mirrors/gpt4free
1
Fork 0
mirror of https://github.com/xtekky/gpt4free.git synced 2025-12-06 02:30:41 -08:00
Code Issues Projects Releases Packages Wiki Activity

Add new providers and enhance existing provider configurations (#2805)

Browse source 
* New provider added(g4f/Provider/Websim.py)

* New provider added(g4f/Provider/Dynaspark.py)

* feat(g4f/gui/client/static/js/chat.v1.js): Enhance provider labeling for HuggingFace integrations

* feat(g4f/gui/server/api.py): add Hugging Face Space compatibility flag to provider data

* feat(g4f/models.py): add new providers and update model configurations

* Update g4f/Provider/__init__.py

* feat(g4f/Provider/AllenAI.py): expand model alias mappings for AllenAI provider

* feat(g4f/Provider/Blackbox.py): restructure image model handling and response processing

* feat(g4f/Provider/PollinationsAI.py): add new model aliases and streamline headers

* Update g4f/Provider/hf_space/*

* refactor(g4f/Provider/Copilot.py): update model alias mapping

* chore(g4f/models.py): update provider configurations for OpenAI models

* docs(docs/providers-and-models.md): update provider tables and model categorization

* fix(etc/examples/vision_images.py): update model and simplify client configuration

* fix(docs/providers-and-models.md): correct streaming status for GlhfChat provider

* docs(docs/providers-and-models.md): update provider capabilities and model documentation

* fix(models): update provider configurations for Mistral models

* fix(g4f/Provider/Blackbox.py): correct model alias key for Mistral variant

* feat(g4f/Provider/hf_space/CohereForAI_C4AI_Command.py): update supported model versions and aliases (close #2802)

* fix(documentation): correct model names and provider counts (https://github.com/xtekky/gpt4free/pull/2805#issuecomment-2727489835)

* fix(g4f/models.py): correct mistral model configurations

* fix(g4f/Provider/DeepInfraChat.py): correct mixtral-small alias key

* New provider added(g4f/Provider/LambdaChat.py)

* feat(g4f/models.py): add new providers and enhance model configurations

* docs(docs/providers-and-models.md): add LambdaChat provider and update model listings

* feat(g4f/models.py): add new Liquid AI model and enhance providers

* docs(docs/providers-and-models.md): update model listings and provider counts

* feat(g4f/Provider/LambdaChat.py): add conditional reasoning processing based on model

* fix(g4f/tools/run_tools.py): handle combined thinking tags in single chunk

* New provider added(g4f/Provider/Goabror.py)

* feat(g4f/Provider/Blackbox.py): implement dynamic session management and model access control

* refactor(g4f/models.py): update provider configurations and model entries

* docs(docs/providers-and-models.md): update model listings and provider counts

---------

Co-authored-by: kqlio67 <>
This commit is contained in:
kqlio67 2025-03-20 16:36:00 +00:00 • committed by GitHub
parent 9c94f9b2e9
commit 52ecfb5019
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
28 changed files with 1183 additions and 306 deletions
Download patch file Download diff file Expand all files Collapse all files

202
docs/providers-and-models.md
View file

@ -17,8 +17,8 @@ This document provides an overview of various AI providers and models, including
- [MiniMax](#providers-minimax)
- [Needs auth](#providers-needs-auth)
- [Models](#models)
- [Text Models](#text-models)
- [Image Models](#image-models)
- [Text generation models](#text-generation-models)
- [Image generation models](#image-generation-models)
- [Conclusion and Usage Tips](#conclusion-and-usage-tips)
---
@ -38,112 +38,117 @@ This document provides an overview of various AI providers and models, including
---
### Providers No auth required
| Website | API Credentials | Provider | Text Models | Image Models | Vision (Image Upload) | Stream | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|
|[playground.allenai.org](https://playground.allenai.org)|No auth required|`g4f.Provider.AllenAI`|`tulu-3-405b, olmo-2-13b, tulu-3-1-8b, tulu-3-70b, olmoe-0125`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[blackbox.ai](https://www.blackbox.ai)|No auth required|`g4f.Provider.Blackbox`|`blackboxai, gemini-1.5-flash, blackboxai-pro, llama-3.1-8b, llama-3.1-70b, llama-3-1-405b, llama-3.3-70b, mixtral-small-28b, deepseek-chat, dbrx-instruct, qwq-32b, hermes-2-dpo, deepseek-v3, deepseek-r1, gemini-2.0-flash, gpt-4o, o1, o3-mini, gemini-1.5-pro, claude-3.7-sonnet` _**(+29)**_|`flux`|`blackboxai, gpt-4o, o1, o3-mini, gemini-1.5-pro, gemini-1.5-flash, llama-3.1-8b, llama-3.1-70b, llama-3.1-405b, gemini-2.0-flash, deepseek-v3`|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chatglm.cn](https://chatglm.cn)|No auth required|`g4f.Provider.ChatGLM`|`glm-4`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chatgpt.com](https://chatgpt.com)|No auth required|`g4f.Provider.ChatGpt`|✔ _**(+7)**_|❌|❌|✔|![Error](https://img.shields.io/badge/HTTPError-f48d37)|
|[chatgpt.es](https://chatgpt.es)|No auth required|`g4f.Provider.ChatGptEs`|`gpt-4, gpt-4o, gpt-4o-mini`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[playground.ai.cloudflare.com](https://playground.ai.cloudflare.com)|[Automatic cookies](https://playground.ai.cloudflare.com)|`g4f.Provider.Cloudflare`|`llama-2-7b, llama-3-8b, llama-3.1-8b, llama-3.2-1b, qwen-1.5-7b`|❌|❌|✔|![Error](https://img.shields.io/badge/Active-brightgreen)|❌|
|[copilot.microsoft.com](https://copilot.microsoft.com)|Optional API key|`g4f.Provider.Copilot`|`gpt-4, gpt-4o`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[duckduckgo.com/aichat](https://duckduckgo.com/aichat)|No auth required|`g4f.Provider.DDG`|`gpt-4, gpt-4o-mini, llama-3.3-70b, claude-3-haiku, o3-mini, mixtral-small-24b`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[deepinfra.com/chat](https://deepinfra.com/chat)|No auth required|`g4f.Provider.DeepInfraChat`|`llama-3.1-8b, llama-3.2-90b, llama-3.3-70b, deepseek-v3, mixtral-small-28b, deepseek-r1, phi-4, wizardlm-2-8x22b, qwen-2.5-72b, yi-34b, qwen-2-72b, dolphin-2.6, dolphin-2.9, dbrx-instruct, airoboros-70b, lzlv-70b, wizardlm-2-7b, mixtral-8x22b, minicpm-2.5`|❌|`llama-3.2-90b, minicpm-2.5`|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chat10.free2gpt.xyz](https://chat10.free2gpt.xyz)|No auth required|`g4f.Provider.Free2GPT`|`gemini-1.5-pro, gemini-1.5-flash`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[freegptsnav.aifree.site](https://freegptsnav.aifree.site)|No auth required|`g4f.Provider.FreeGpt`|`gemini-1.5-pro, gemini-1.5-flash`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[app.giz.ai/assistant](https://app.giz.ai/assistant)|No auth required|`g4f.Provider.GizAI`|`gemini-1.5-flash`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[glider.so](https://glider.so)|No auth required|`g4f.Provider.Glider`|`llama-3.1-70b, llama-3.1-8b, llama-3.2-3b, deepseek-r1`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[hailuo.ai](https://www.hailuo.ai)|No auth required|`g4f.Provider.HailuoAI`|`MiniMax` _**(1)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[editor.imagelabs.net](editor.imagelabs.net)|No auth required|`g4f.Provider.ImageLabs`|❌|`sdxl-turbo`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[huggingface.co/spaces](https://huggingface.co/spaces)|Optional API key|`g4f.Provider.HuggingSpace`|`qvq-72b, qwen-2-72b, command-r, command-r-plus, command-r7b`|`flux-dev, flux-schnell, sd-3.5`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[jmuz.me](https://jmuz.me)|Optional API key|`g4f.Provider.Jmuz`|`claude-3-haiku, claude-3-opus, claude-3-haiku, claude-3.5-sonnet, deepseek-r1, deepseek-chat, gemini-exp, gemini-1.5-flash, gemini-1.5-pro, gemini-2.0-flash-thinking, gpt-4, gpt-4o, gpt-4o-mini, llama-3-70b, llama-3-8b, llama-3.1-405b, llama-3.1-70b, llama-3.1-8b, llama-3.2-11b, llama-3.2-90b, llama-3.3-70b, mixtral-8x7b, qwen-2.5-72b, qwen-2.5-coder-32b, qwq-32b, wizardlm-2-8x22b`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[liaobots.work](https://liaobots.work)|[Automatic cookies](https://liaobots.work)|`g4f.Provider.Liaobots`|`claude-3.5-sonnet, claude-3.7-sonnet, claude-3.7-sonnet-thinking, claude-3-opus, claude-3-sonnet, deepseek-r1, deepseek-v3, gemini-2.0-flash, gemini-2.0-flash-thinking, gemini-2.0-pro, gpt-4, gpt-4o, gpt-4o-mini, grok-3, grok-3-r1, o3-mini`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[oi-vscode-server.onrender.com](https://oi-vscode-server.onrender.com)|No auth required|`g4f.Provider.OIVSCode`|`gpt-4o-mini, deepseek-v3`|❌|`gpt-4o-mini`|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[labs.perplexity.ai](https://labs.perplexity.ai)|No auth required|`g4f.Provider.PerplexityLabs`|`sonar, sonar-pro, sonar-reasoning, sonar-reasoning-pro`|❌|❌|✔|![Error](https://img.shields.io/badge/Active-brightgreen)|
|[pi.ai/talk](https://pi.ai/talk)|[Manual cookies](https://pi.ai/talk)|`g4f.Provider.Pi`|`pi`|❌|❌|✔|![Error](https://img.shields.io/badge/Active-brightgreen)|
|[pizzagpt.it](https://www.pizzagpt.it)|No auth required|`g4f.Provider.Pizzagpt`|`gpt-4o-mini`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[pollinations.ai](https://pollinations.ai)|No auth required|`g4f.Provider.PollinationsAI`|`gpt-4o-mini, gpt-4o, o1-mini, qwen-2.5-coder-32b, llama-3.3-70b, mistral-nemo, llama-3.1-8b, deepseek-r1, phi-4` _**(6+)**_|`flux, flux-pro, flux-dev, flux-schnell, dall-e-3, sdxl-turbo`|gpt-4o, gpt-4o-mini, o1-mini|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[pollinations.ai](https://pollinations.ai)|No auth required|`g4f.Provider.PollinationsImage`|❌|`flux, flux-pro, flux-dev, flux-schnell, dall-e-3, sdxl-turbo`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[app.prodia.com](https://app.prodia.com)|No auth required|`g4f.Provider.Prodia`|❌|✔ _**(46)**_|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[teach-anything.com](https://www.teach-anything.com)|No auth required|`g4f.Provider.TeachAnything`|`gemini-1.5-pro, gemini-1.5-flash`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[you.com](https://you.com)|[Manual cookies](https://you.com)|`g4f.Provider.You`|✔|✔|✔|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chat9.yqcloud.top](https://chat9.yqcloud.top)|No auth required|`g4f.Provider.Yqcloud`|`gpt-4`|✔|✔|✔|![](https://img.shields.io/badge/Active-brightgreen)|
| Website | API Credentials | Provider | Text generation | Image generation | Audio generation | Vision (Image Upload) | Streaming | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|------|
|[playground.allenai.org](https://playground.allenai.org)|No auth required|`g4f.Provider.AllenAI`|`tulu-3-405b, olmo-2-13b, tulu-3-1-8b, tulu-3-70b, olmoe-0125`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[ai-arta.com](https://ai-arta.com)|No auth required|`g4f.Provider.ARTA`|❌|✔ _**(17+)**_|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[blackbox.ai](https://www.blackbox.ai)|No auth required|`g4f.Provider.Blackbox`|`blackboxai, gpt-4o-mini, deepseek-chat, deepseek-v3, deepseek-r1, gpt-4o, o1, o3-mini, claude-3.7-sonnet` _**(40+)**_|`flux`|❌|`blackboxai, gpt-4o, o1, o3-mini, deepseek-v3` _**(7+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chatglm.cn](https://chatglm.cn)|No auth required|`g4f.Provider.ChatGLM`|`glm-4`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chatgpt.com](https://chatgpt.com)|No auth required|`g4f.Provider.ChatGpt`|✔ _**(7+)**_|❌|❌|❌|✔|![Error](https://img.shields.io/badge/HTTPError-f48d37)|
|[chatgpt.es](https://chatgpt.es)|No auth required|`g4f.Provider.ChatGptEs`|`gpt-4, gpt-4o, gpt-4o-mini`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[playground.ai.cloudflare.com](https://playground.ai.cloudflare.com)|[Automatic cookies](https://playground.ai.cloudflare.com)|`g4f.Provider.Cloudflare`|`llama-2-7b, llama-3-8b, llama-3.1-8b, llama-3.2-1b, qwen-1.5-7b`|❌|❌|❌|✔|![Error](https://img.shields.io/badge/Active-brightgreen)|❌|
|[copilot.microsoft.com](https://copilot.microsoft.com)|Optional API key|`g4f.Provider.Copilot`|`gpt-4, o1`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[duckduckgo.com/aichat](https://duckduckgo.com/aichat)|No auth required|`g4f.Provider.DDG`|`gpt-4, gpt-4o-mini, llama-3.3-70b, claude-3-haiku, o3-mini, mixtral-small-24b`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[deepinfra.com/chat](https://deepinfra.com/chat)|No auth required|`g4f.Provider.DeepInfraChat`|`llama-3.1-8b, llama-3.2-90b, llama-3.3-70b, deepseek-v3, mixtral-small-24b, deepseek-r1, phi-4, wizardlm-2-8x22b, qwen-2.5-72b, yi-34b, qwen-2-72b, dolphin-2.6, dolphin-2.9, dbrx-instruct, airoboros-70b, lzlv-70b, wizardlm-2-7b, mixtral-8x22b, minicpm-2.5`|❌|❌|`llama-3.2-90b, minicpm-2.5`|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[dynaspark.onrender.com](https://dynaspark.onrender.com)|No auth required|`g4f.Provider.Dynaspark`|`gemini-1.5-flash, gemini-2.0-flash`|❌|❌|`gemini-1.5-flash, gemini-2.0-flash`|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chat10.free2gpt.xyz](https://chat10.free2gpt.xyz)|No auth required|`g4f.Provider.Free2GPT`|`gemini-1.5-pro, gemini-1.5-flash`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[freegptsnav.aifree.site](https://freegptsnav.aifree.site)|No auth required|`g4f.Provider.FreeGpt`|`gemini-1.5-pro, gemini-1.5-flash`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[app.giz.ai/assistant](https://app.giz.ai/assistant)|No auth required|`g4f.Provider.GizAI`|`gemini-1.5-flash`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[glider.so](https://glider.so)|No auth required|`g4f.Provider.Glider`|`llama-3.1-70b, llama-3.1-8b, llama-3.2-3b, deepseek-r1`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[goabror.uz](https://goabror.uz)|No auth required|`g4f.Provider.Goabror`|`gpt-4`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[hailuo.ai](https://www.hailuo.ai)|No auth required|`g4f.Provider.HailuoAI`|`MiniMax` _**(1+)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[editor.imagelabs.net](editor.imagelabs.net)|No auth required|`g4f.Provider.ImageLabs`|❌|`sdxl-turbo`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[huggingface.co/spaces](https://huggingface.co/spaces)|Optional API key|`g4f.Provider.HuggingSpace`|`qvq-72b, qwen-2-72b, command-r, command-r-plus, command-r7b, command-a`|`flux-dev, flux-schnell, sd-3.5`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[jmuz.me](https://jmuz.me)|Optional API key|`g4f.Provider.Jmuz`|`claude-3-haiku, claude-3-opus, claude-3.5-sonnet, deepseek-r1, deepseek-chat, gemini-exp, gemini-1.5-flash, gemini-1.5-pro, gemini-2.0-flash-thinking, gpt-4, gpt-4o, gpt-4o-mini, llama-3-70b, llama-3-8b, llama-3.1-405b, llama-3.1-70b, llama-3.1-8b, llama-3.2-11b, llama-3.2-90b, llama-3.3-70b, mixtral-8x7b, qwen-2.5-72b, qwen-2.5-coder-32b, qwq-32b, wizardlm-2-8x22b`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[lambda.chat](https://lambda.chat)|No auth required|`g4f.Provider.LambdaChat`|`deepseek-v3, deepseek-r1, hermes-3, nemotron-70b, llama-3.3-70b`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[liaobots.work](https://liaobots.work)|[Automatic cookies](https://liaobots.work)|`g4f.Provider.Liaobots`|`claude-3.5-sonnet, claude-3.7-sonnet, claude-3.7-sonnet-thinking, claude-3-opus, claude-3-sonnet, deepseek-r1, deepseek-v3, gemini-2.0-flash, gemini-2.0-flash-thinking, gemini-2.0-pro, gpt-4, gpt-4o, gpt-4o-mini, grok-3, grok-3-r1, o3-mini`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[oi-vscode-server.onrender.com](https://oi-vscode-server.onrender.com)|No auth required|`g4f.Provider.OIVSCode`|`gpt-4o-mini, deepseek-v3`|❌|❌|`gpt-4o-mini`|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[labs.perplexity.ai](https://labs.perplexity.ai)|No auth required|`g4f.Provider.PerplexityLabs`|`sonar, sonar-pro, sonar-reasoning, sonar-reasoning-pro`|❌|❌|❌|✔|![Error](https://img.shields.io/badge/Active-brightgreen)|
|[pi.ai/talk](https://pi.ai/talk)|[Manual cookies](https://pi.ai/talk)|`g4f.Provider.Pi`|`pi`|❌|❌|❌|✔|![Error](https://img.shields.io/badge/Active-brightgreen)|
|[pizzagpt.it](https://www.pizzagpt.it)|No auth required|`g4f.Provider.Pizzagpt`|`gpt-4o-mini`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[pollinations.ai](https://pollinations.ai)|No auth required|`g4f.Provider.PollinationsAI`|`gpt-4o-mini, gpt-4o, o1-mini, qwen-2.5-coder-32b, llama-3.3-70b, mistral-nemo, llama-3.1-8b, deepseek-r1, phi-4` _**(9+)**_|`flux, flux-pro, flux-dev, flux-schnell, dall-e-3, sdxl-turbo`|✔ _**(1+)**_|`gpt-4o, gpt-4o-mini, o1-mini, o3-mini`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[pollinations.ai](https://pollinations.ai)|No auth required|`g4f.Provider.PollinationsImage`|❌|`flux, flux-pro, flux-dev, flux-schnell, dall-e-3, sdxl-turbo`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[teach-anything.com](https://www.teach-anything.com)|No auth required|`g4f.Provider.TeachAnything`|`gemini-1.5-pro, gemini-1.5-flash`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[you.com](https://you.com)|[Manual cookies](https://you.com)|`g4f.Provider.You`|✔|✔|❌|✔|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[websim.ai](https://websim.ai)|No auth required|`g4f.Provider.Websim`|`gemini-1.5-pro, gemini-1.5-flash`|`flux`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chat9.yqcloud.top](https://chat9.yqcloud.top)|No auth required|`g4f.Provider.Yqcloud`|`gpt-4`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
---
### Providers HuggingFace
| Website | API Credentials | Provider | Text Models | Image Models | Vision (Image Upload) | Stream | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|
|[huggingface.co/chat](https://huggingface.co/chat)|[Manual cookies](https://huggingface.co/chat)|`g4f.Provider.HuggingChat`|`qwen-2.5-72b, llama-3.3-70b, command-r-plus, deepseek-r1, qwq-32b, nemotron-70b, llama-3.2-11b, mistral-nemo, phi-3.5-mini`|`flux-dev, flux-schnell`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[huggingface.co/chat](https://huggingface.co/chat)|[API key / Cookies](https://huggingface.co/settings/tokens)|`g4f.Provider.HuggingFace`|✔ _**(47+)**_|✔ _**(9+)**_|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[api-inference.huggingface.co](https://api-inference.huggingface.co)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.HuggingFaceAPI`|✔ _**(9+)**_|✔ _**(2+)**_|✔ _**(1+)**_|❌|![](https://img.shields.io/badge/Active-brightgreen)|✔|
| Website | API Credentials | Provider | Text generation | Image generation | Audio generation | Vision (Image Upload) | Streaming | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|------|
|[huggingface.co/chat](https://huggingface.co/chat)|[Manual cookies](https://huggingface.co/chat)|`g4f.Provider.HuggingChat`|`qwen-2.5-72b, llama-3.3-70b, command-r-plus, deepseek-r1, qwq-32b, nemotron-70b, llama-3.2-11b, mistral-nemo, phi-3.5-mini`|`flux-dev, flux-schnell`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[huggingface.co/chat](https://huggingface.co/chat)|[API key / Cookies](https://huggingface.co/settings/tokens)|`g4f.Provider.HuggingFace`|✔ _**(47+)**_|✔ _**(9+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[api-inference.huggingface.co](https://api-inference.huggingface.co)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.HuggingFaceAPI`|✔ _**(9+)**_|✔ _**(2+)**_|❌|✔ _**(1+)**_|❌|![](https://img.shields.io/badge/Active-brightgreen)|✔|
---
### Providers HuggingSpace
| Website | API Credentials | Provider | Text Models | Image Models | Vision (Image Upload) | Stream | Status | Auth |
| Website | API Credentials | Provider | Text generation | Image generation | Audio generation | Vision (Image Upload) | Streaming | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|------|
|[black-forest-labs-flux-1-dev.hf.space](https://black-forest-labs-flux-1-dev.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.BlackForestLabsFlux1Dev`|❌|`flux-dev`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[black-forest-labs-flux-1-schnell.hf.space](https://black-forest-labs-flux-1-schnell.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.BlackForestLabsFlux1Schnell`|❌|`flux-schnell`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[cohereforai-c4ai-command.hf.space](https://cohereforai-c4ai-command.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.CohereForAI`|`command-r, command-r-plus, command-r7b`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[roxky-flux-1-dev.hf.space](https://roxky-flux-1-dev.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|❌|❌|✔ _**(3)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[huggingface.co/spaces/deepseek-ai/Janus-Pro-7B](https://huggingface.co/spaces/deepseek-ai/Janus-Pro-7B)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Janus_Pro_7B`|✔|✔|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[qwen-qvq-72b-preview.hf.space](https://qwen-qvq-72b-preview.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Qwen_QVQ_72B`|`qvq-72b`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[qwen-qwen2-5-1m-demo.hf.space](https://qwen-qwen2-5-1m-demo.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Qwen_Qwen_2_5M_Demo`|`qwen-2.5-1m-demo`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[qwen-qwen2-72b-instruct.hf.space](https://qwen-qwen2-72b-instruct.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Qwen_Qwen_2_72B_Instruct`|`qwen-2-72b`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[stabilityai-stable-diffusion-3-5-large.hf.space](https://stabilityai-stable-diffusion-3-5-large.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.StableDiffusion35Large`|❌|`sd-3.5`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[voodoohop-flux-1-schnell.hf.space](https://voodoohop-flux-1-schnell.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.VoodoohopFlux1Schnell`|❌|`flux-schnell`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[black-forest-labs-flux-1-dev.hf.space](https://black-forest-labs-flux-1-dev.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.BlackForestLabs_Flux1Dev`|❌|`flux-dev`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[black-forest-labs-flux-1-schnell.hf.space](https://black-forest-labs-flux-1-schnell.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.BlackForestLabs_Flux1Schnell`|❌|`flux-schnell`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[cohereforai-c4ai-command.hf.space](https://cohereforai-c4ai-command.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.CohereForAI_C4AI_Command`|`command-r, command-r-plus, command-r7b`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[huggingface.co/spaces/deepseek-ai/Janus-Pro-7B](https://huggingface.co/spaces/deepseek-ai/Janus-Pro-7B)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.DeepseekAI_Janus_Pro_7b`|✔|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[roxky-flux-1-dev.hf.space](https://roxky-flux-1-dev.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.G4F`|✔ _**(1+)**_|✔ _**(4+)**_|❌|✔ _**(1+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[microsoft-phi-4-multimodal.hf.space](https://microsoft-phi-4-multimodal.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Microsoft_Phi_4`|✔ _**(1+)**_|❌|❌|✔ _**(1+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[qwen-qvq-72b-preview.hf.space](https://qwen-qvq-72b-preview.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Qwen_QVQ_72B`|`qvq-72b`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[qwen-qwen2-5-1m-demo.hf.space](https://qwen-qwen2-5-1m-demo.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Qwen_Qwen_2_5M_Demo`|`qwen-2.5-1m-demo`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[qwen-qwen2-72b-instruct.hf.space](https://qwen-qwen2-72b-instruct.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Qwen_Qwen_2_72B_Instruct`|`qwen-2-72b`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[stabilityai-stable-diffusion-3-5-large.hf.space](https://stabilityai-stable-diffusion-3-5-large.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.StabilityAI_SD35Large`|❌|`sd-3.5`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[voodoohop-flux-1-schnell.hf.space](https://voodoohop-flux-1-schnell.hf.space)|[Get API key](https://huggingface.co/settings/tokens)|`g4f.Provider.Voodoohop_Flux1Schnell`|❌|`flux-schnell`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
### Providers Local
| Website | API Credentials | Provider | Text Models | Image Models | Vision (Image Upload) | Stream | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|
|[]( )|No auth required|`g4f.Provider.Local`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[ollama.com](https://ollama.com)|No auth required|`g4f.Provider.Ollama`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
| Website | API Credentials | Provider | Text generation | Image generation | Audio generation | Vision (Image Upload) | Streaming | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|------|
|[]( )|No auth required|`g4f.Provider.Local`|✔|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[ollama.com](https://ollama.com)|No auth required|`g4f.Provider.Ollama`|✔|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
---
### Providers MiniMax
| Website | API Credentials | Provider | Text Models | Image Models | Vision (Image Upload) | Stream | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|
|[hailuo.ai/chat](https://www.hailuo.ai/chat)|[Get API key](https://intl.minimaxi.com/user-center/basic-information/interface-key)|`g4f.Provider.MiniMax`|`MiniMax` _**(1)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
| Website | API Credentials | Provider | Text generation | Image generation| Audio generation | Vision (Image Upload) | Stream | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|------|
|[hailuo.ai/chat](https://www.hailuo.ai/chat)|[Get API key](https://intl.minimaxi.com/user-center/basic-information/interface-key)|`g4f.Provider.MiniMax`|`MiniMax` _**(1)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
---
### Providers Needs Auth
| Website | API Credentials | Provider | Text Models | Image Models | Vision (Image Upload) | Stream | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|
|[console.anthropic.com](https://console.anthropic.com)|[Get API key](https://console.anthropic.com/settings/keys)|`g4f.Provider.Anthropic`|✔ _**(8+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[bing.com/images/create](https://www.bing.com/images/create)|[Manual cookies](https://www.bing.com)|`g4f.Provider.BingCreateImages`|❌|`dall-e-3`|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[cablyai.com/chat](https://cablyai.com/chat)|[Get API key](https://cablyai.com)|`g4f.Provider.CablyAI`|✔|✔|✔|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[inference.cerebras.ai](https://inference.cerebras.ai/)|[Get API key](https://cloud.cerebras.ai)|`g4f.Provider.Cerebras`|✔ _**(3+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[copilot.microsoft.com](https://copilot.microsoft.com)|[Manual cookies](https://copilot.microsoft.com)|`g4f.Provider.CopilotAccount`|✔ _**(1+)**_|✔ _**(1+)**_|✔ _**(1+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[deepinfra.com](https://deepinfra.com)|[Get API key](https://deepinfra.com/dash/api_keys)|`g4f.Provider.DeepInfra`|✔ _**(17+)**_|✔ _**(6+)**_|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[platform.deepseek.com](https://platform.deepseek.com)|[Get API key](https://platform.deepseek.com/api_keys)|`g4f.Provider.DeepSeek`|✔ _**(1+)**_|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[gemini.google.com](https://gemini.google.com)|[Manual cookies](https://gemini.google.com)|`g4f.Provider.Gemini`|`gemini-2.0`|`gemini-2.0`|`gemini-2.0`|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[ai.google.dev](https://ai.google.dev)|[Get API key](https://aistudio.google.com/u/0/apikey)|`g4f.Provider.GeminiPro`|`gemini-1.5-flash, gemini-1.5-pro, gemini-2.0-flash`|❌|`gemini-1.5-pro`|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[developers.sber.ru/gigachat](https://developers.sber.ru/gigachat)|[Manual cookies](https://developers.sber.ru/gigachat)|`g4f.Provider.GigaChat`|✔ _**(3+)**_|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[github.com/copilot](https://github.com/copilot)|[Manual cookies](https://github.com/copilot)|`g4f.Provider.GithubCopilot`|✔ _**(4+)**_|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[glhf.chat](https://glhf.chat)|[Get API key](https://glhf.chat/user-settings/api)|`g4f.Provider.GlhfChat`|✔ _**(22+)**_|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[console.groq.com/playground](https://console.groq.com/playground)|[Get API key](https://console.groq.com/keys)|`g4f.Provider.Groq`|✔ _**(18+)**_|❌|✔|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[meta.ai](https://www.meta.ai)|[Manual cookies](https://www.meta.ai)|`g4f.Provider.MetaAI`|`meta-ai`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|✔|
|[meta.ai](https://www.meta.ai)|[Manual cookies](https://www.meta.ai)|`g4f.Provider.MetaAIAccount`|❌|`meta-ai`|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[designer.microsoft.com](https://designer.microsoft.com)|[Manual cookies](https://designer.microsoft.com)|`g4f.Provider.MicrosoftDesigner`|❌|`dall-e-3`|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[platform.openai.com](https://platform.openai.com)|[Get API key](https://platform.openai.com/settings/organization/api-keys)|`g4f.Provider.OpenaiAPI`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chatgpt.com](https://chatgpt.com)|[Manual cookies](https://chatgpt.com)|`g4f.Provider.OpenaiChat`|`gpt-4o, gpt-4o-mini, gpt-4` _**(8+)**_|✔_**(1)**_|✔_**(8+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[perplexity.ai](https://www.perplexity.ai)|[Get API key](https://www.perplexity.ai/settings/api)|`g4f.Provider.PerplexityApi`|✔ _**(6+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chat.reka.ai](https://chat.reka.ai)|[Manual cookies](https://chat.reka.ai)|`g4f.Provider.Reka`|`reka-core`|❌|✔|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[replicate.com](https://replicate.com)|[Get API key](https://replicate.com/account/api-tokens)|`g4f.Provider.Replicate`|✔ _**(1+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[beta.theb.ai](https://beta.theb.ai)|[Get API key](https://beta.theb.ai)|`g4f.Provider.ThebApi`|✔ _**(21+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[whiterabbitneo.com](https://www.whiterabbitneo.com)|[Manual cookies](https://www.whiterabbitneo.com)|`g4f.Provider.WhiteRabbitNeo`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[console.x.ai](https://console.x.ai)|[Get API key](https://console.x.ai)|`g4f.Provider.xAI`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
| Website | API Credentials | Provider | Text generation | Image generation | Audio generation | Vision (Image Upload) | Streaming | Status |
|----------|-------------|--------------|---------------|--------|--------|------|------|------|
|[console.anthropic.com](https://console.anthropic.com)|[Get API key](https://console.anthropic.com/settings/keys)|`g4f.Provider.Anthropic`|✔ _**(8+)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[bing.com/images/create](https://www.bing.com/images/create)|[Manual cookies](https://www.bing.com)|`g4f.Provider.BingCreateImages`|❌|`dall-e-3`|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[cablyai.com/chat](https://cablyai.com/chat)|[Get API key](https://cablyai.com)|`g4f.Provider.CablyAI`|✔|✔|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[inference.cerebras.ai](https://inference.cerebras.ai/)|[Get API key](https://cloud.cerebras.ai)|`g4f.Provider.Cerebras`|✔ _**(3+)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[copilot.microsoft.com](https://copilot.microsoft.com)|[Manual cookies](https://copilot.microsoft.com)|`g4f.Provider.CopilotAccount`|✔ _**(1+)**_|✔ _**(1+)**_|❌|✔ _**(1+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[deepinfra.com](https://deepinfra.com)|[Get API key](https://deepinfra.com/dash/api_keys)|`g4f.Provider.DeepInfra`|✔ _**(17+)**_|✔ _**(6+)**_|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[platform.deepseek.com](https://platform.deepseek.com)|[Get API key](https://platform.deepseek.com/api_keys)|`g4f.Provider.DeepSeek`|✔ _**(1+)**_|❌|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[gemini.google.com](https://gemini.google.com)|[Manual cookies](https://gemini.google.com)|`g4f.Provider.Gemini`|`gemini-2.0`|`gemini-2.0`|❌|`gemini-2.0`|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[ai.google.dev](https://ai.google.dev)|[Get API key](https://aistudio.google.com/u/0/apikey)|`g4f.Provider.GeminiPro`|`gemini-1.5-flash, gemini-1.5-pro, gemini-2.0-flash`|❌|❌|`gemini-1.5-pro`|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[developers.sber.ru/gigachat](https://developers.sber.ru/gigachat)|[Manual cookies](https://developers.sber.ru/gigachat)|`g4f.Provider.GigaChat`|✔ _**(3+)**_|❌|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[github.com/copilot](https://github.com/copilot)|[Manual cookies](https://github.com/copilot)|`g4f.Provider.GithubCopilot`|✔ _**(4+)**_|❌|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[glhf.chat](https://glhf.chat)|[Get API key](https://glhf.chat/user-settings/api)|`g4f.Provider.GlhfChat`|✔ _**(22+)**_|❌|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[console.groq.com/playground](https://console.groq.com/playground)|[Get API key](https://console.groq.com/keys)|`g4f.Provider.Groq`|✔ _**(18+)**_|❌|❌|✔|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[meta.ai](https://www.meta.ai)|[Manual cookies](https://www.meta.ai)|`g4f.Provider.MetaAI`|`meta-ai`|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|✔|
|[meta.ai](https://www.meta.ai)|[Manual cookies](https://www.meta.ai)|`g4f.Provider.MetaAIAccount`|❌|`meta-ai`|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[designer.microsoft.com](https://designer.microsoft.com)|[Manual cookies](https://designer.microsoft.com)|`g4f.Provider.MicrosoftDesigner`|❌|`dall-e-3`|❌|❌|❌|![](https://img.shields.io/badge/Active-brightgreen)|
|[platform.openai.com](https://platform.openai.com)|[Get API key](https://platform.openai.com/settings/organization/api-keys)|`g4f.Provider.OpenaiAPI`|✔|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chatgpt.com](https://chatgpt.com)|[Manual cookies](https://chatgpt.com)|`g4f.Provider.OpenaiChat`|`gpt-4o, gpt-4o-mini, gpt-4` _**(8+)**_|✔ _**(1)**_|❌|✔ _**(8+)**_|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[perplexity.ai](https://www.perplexity.ai)|[Get API key](https://www.perplexity.ai/settings/api)|`g4f.Provider.PerplexityApi`|✔ _**(6+)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[chat.reka.ai](https://chat.reka.ai)|[Manual cookies](https://chat.reka.ai)|`g4f.Provider.Reka`|`reka-core`|✔|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[replicate.com](https://replicate.com)|[Get API key](https://replicate.com/account/api-tokens)|`g4f.Provider.Replicate`|✔ _**(1+)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[beta.theb.ai](https://beta.theb.ai)|[Get API key](https://beta.theb.ai)|`g4f.Provider.ThebApi`|✔ _**(21+)**_|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[whiterabbitneo.com](https://www.whiterabbitneo.com)|[Manual cookies](https://www.whiterabbitneo.com)|`g4f.Provider.WhiteRabbitNeo`|✔|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
|[console.x.ai](https://console.x.ai)|[Get API key](https://console.x.ai)|`g4f.Provider.xAI`|✔|❌|❌|❌|✔|![](https://img.shields.io/badge/Active-brightgreen)|
---
## Models
### Text Models
### Text generation models
| Model | Base Provider | Providers | Website |
|-------|---------------|-----------|---------|
|gpt-4|OpenAI|8+ Providers|[platform.openai.com](https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4)|
|gpt-4o|OpenAI|7+ Providers|[platform.openai.com](https://platform.openai.com/docs/models/gpt-4o)|
|gpt-4o-mini|OpenAI|7+ Providers|[platform.openai.com](https://platform.openai.com/docs/models/gpt-4o-mini)|
|o1|OpenAI|2+ Providers|[openai.com](https://openai.com/index/introducing-openai-o1-preview/)|
|gpt-4|OpenAI|9+ Providers|[platform.openai.com](https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4)|
|gpt-4o|OpenAI|6+ Providers|[platform.openai.com](https://platform.openai.com/docs/models/gpt-4o)|
|gpt-4o-mini|OpenAI|8+ Providers|[platform.openai.com](https://platform.openai.com/docs/models/gpt-4o-mini)|
|o1|OpenAI|3+ Providers|[openai.com](https://openai.com/index/introducing-openai-o1-preview/)|
|o1-mini|OpenAI|1+ Providers|[openai.com](https://openai.com/index/openai-o1-mini-advancing-cost-efficient-reasoning/)|
|o3-mini|OpenAI|3+ Providers|[openai.com](https://openai.com/index/openai-o3-mini/)|
|gigachat|GigaChat|1+ Providers|[developers.sber.ru/gigachat](https://developers.sber.ru/gigachat)|
@ -162,17 +167,16 @@ This document provides an overview of various AI providers and models, including
|mixtral-8x7b|Mistral|1+ Providers|[mistral.ai](https://mistral.ai/news/mixtral-of-experts/)|
|mixtral-8x22b|Mistral|1+ Providers|[huggingface.co](https://huggingface.co/mistralai/Mixtral-8x22B-Instruct-v0.1)|
|mistral-nemo|Mistral|3+ Providers|[huggingface.co](https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407)|
|mixtral-small-24b|Mistral|1+ Providers|[huggingface.co](https://huggingface.co/mistralai/Mistral-Small-24B-Instruct-2501)|
|mixtral-small-28b|Mistral|2+ Providers|[mistral.ai](https://mistral.ai/news/mixtral-small-28b/)|
|hermes-2-dpo|NousResearch|1+ Providers|[huggingface.co](https://huggingface.co/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO)|
|mixtral-small-24b|Mistral|2+ Providers|[huggingface.co](https://huggingface.co/mistralai/Mistral-Small-24B-Instruct-2501)|
|hermes-3|NousResearch|1+ Providers|[huggingface.co](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-405B-FP8)|
|phi-3.5-mini|Microsoft|1+ Providers|[huggingface.co](https://huggingface.co/microsoft/Phi-3.5-mini-instruct)|
|phi-4|Microsoft|2+ Providers|[techcommunity.microsoft.com](https://techcommunity.microsoft.com/blog/aiplatformblog/introducing-phi-4-microsoft%E2%80%99s-newest-small-language-model-specializing-in-comple/4357090)|
|wizardlm-2-7b|Microsoft|1+ Providers|[wizardlm.github.io](https://wizardlm.github.io/WizardLM2/)|
|wizardlm-2-8x22b|Microsoft|2+ Providers|[wizardlm.github.io](https://wizardlm.github.io/WizardLM2/)|
|gemini-2.0|Google DeepMind|1+ Providers|[deepmind.google](http://deepmind.google/technologies/gemini/)|
|gemini-exp|Google DeepMind|1+ Providers|[blog.google](https://blog.google/feed/gemini-exp-1206/)|
|gemini-1.5-flash|Google DeepMind|6+ Providers|[deepmind.google](https://deepmind.google/technologies/gemini/flash/)|
|gemini-1.5-flash|Google DeepMind|7+ Providers|[deepmind.google](https://deepmind.google/technologies/gemini/flash/)|
|gemini-1.5-pro|Google DeepMind|6+ Providers|[deepmind.google](https://deepmind.google/technologies/gemini/pro/)|
|gemini-2.0|Google DeepMind|1+ Providers|[deepmind.google](http://deepmind.google/technologies/gemini/)|
|gemini-2.0-flash|Google DeepMind|3+ Providers|[deepmind.google](https://deepmind.google/technologies/gemini/flash/)|
|gemini-2.0-flash-thinking|Google DeepMind|1+ Providers|[ai.google.dev](https://ai.google.dev/gemini-api/docs/thinking-mode)|
|gemini-2.0-pro|Google DeepMind|1+ Providers|[deepmind.google](https://deepmind.google/technologies/gemini/flash-thinking/)|
@ -184,22 +188,22 @@ This document provides an overview of various AI providers and models, including
|claude-3.7-sonnet-thinking|Anthropic|1+ Providers|[anthropic.com](https://www.anthropic.com/claude/sonnet)|
|reka-core|Reka AI|1+ Providers|[reka.ai](https://www.reka.ai/ourmodels)|
|blackboxai|Blackbox AI|1+ Providers|[docs.blackbox.chat](https://docs.blackbox.chat/blackbox-ai-1)|
|blackboxai-pro|Blackbox AI|1+ Providers|[docs.blackbox.chat](https://docs.blackbox.chat/blackbox-ai-1)|
|command-r|CohereForAI|1+ Providers|[docs.cohere.com](https://docs.cohere.com/docs/command-r-plus)|
|command-r-plus|CohereForAI|2+ Providers|[docs.cohere.com](https://docs.cohere.com/docs/command-r-plus)|
|command-r7b|CohereForAI|1+ Providers|[huggingface.co](https://huggingface.co/CohereForAI/c4ai-command-r7b-12-2024)|
|command-a|CohereForAI|1+ Providers|[docs.cohere.com](https://docs.cohere.com/v2/docs/command-a)|
|qwen-1.5-7b|Qwen|1+ Providers|[huggingface.co](https://huggingface.co/Qwen/Qwen1.5-7B)|
|qwen-2-72b|Qwen|2+ Providers|[huggingface.co](https://huggingface.co/Qwen/Qwen2-72B)|
|qwen-2-vl-7b|Qwen|1+ Providers|[huggingface.co](https://huggingface.co/Qwen/Qwen2-VL-7B)|
|qwen-2.5-72b|Qwen|1+ Providers|[huggingface.co](https://huggingface.co/Qwen/Qwen2.5-72B-Instruct)|
|qwen-2.5-coder-32b|Qwen|3+ Providers|[huggingface.co](https://huggingface.co/Qwen/Qwen2.5-Coder-32B)|
|qwen-2.5-1m-demo|Qwen|1+ Providers|[huggingface.co](https://huggingface.co/Qwen/Qwen2.5-1M-Demo)|
|qwq-32b|Qwen|3+ Providers|[huggingface.co](https://huggingface.co/Qwen/QwQ-32B-Preview)|
|qwq-32b|Qwen|2+ Providers|[huggingface.co](https://huggingface.co/Qwen/QwQ-32B-Preview)|
|qvq-72b|Qwen|1+ Providers|[huggingface.co](https://huggingface.co/Qwen/QVQ-72B-Preview)|
|pi|Inflection|1+ Providers|[inflection.ai](https://inflection.ai/blog/inflection-2-5)|
|deepseek-chat|DeepSeek|2+ Providers|[huggingface.co](https://huggingface.co/deepseek-ai/deepseek-llm-67b-chat)|
|deepseek-v3|DeepSeek|4+ Providers|[api-docs.deepseek.com](https://api-docs.deepseek.com/news/news250120)|
|deepseek-r1|DeepSeek|8+ Providers|[api-docs.deepseek.com](https://api-docs.deepseek.com/news/news250120)|
|deepseek-r1|DeepSeek|9+ Providers|[api-docs.deepseek.com](https://api-docs.deepseek.com/news/news250120)|
|janus-pro-7b|DeepSeek|2+ Providers|[api-docs.deepseek.com](https://api-docs.deepseek.com/docs/janus-pro-7b)|
|grok-3|x.ai|1+ Providers|[x.ai](https://x.ai/blog/grok-3)|
|grok-3-r1|x.ai|1+ Providers|[x.ai](https://x.ai/blog/grok-3)|
@ -208,8 +212,8 @@ This document provides an overview of various AI providers and models, including
|sonar-reasoning|Perplexity AI|1+ Providers|[sonar.perplexity.ai](https://sonar.perplexity.ai/)|
|sonar-reasoning-pro|Perplexity AI|1+ Providers|[sonar.perplexity.ai](https://sonar.perplexity.ai/)|
|r1-1776|Perplexity AI|1+ Providers|[perplexity.ai](https://www.perplexity.ai/hub/blog/open-sourcing-r1-1776)|
|nemotron-70b|Nvidia|2+ Providers|[build.nvidia.com](https://build.nvidia.com/nvidia/llama-3_1-nemotron-70b-instruct)|
|dbrx-instruct|Databricks|2+ Providers|[huggingface.co](https://huggingface.co/databricks/dbrx-instruct)|
|nemotron-70b|Nvidia|3+ Providers|[build.nvidia.com](https://build.nvidia.com/nvidia/llama-3_1-nemotron-70b-instruct)|
|dbrx-instruct|Databricks|1+ Providers|[huggingface.co](https://huggingface.co/databricks/dbrx-instruct)|
|glm-4|THUDM|1+ Providers|[github.com/THUDM](https://github.com/THUDM/GLM-4)|
|mini_max|MiniMax|1+ Providers|[hailuo.ai](https://www.hailuo.ai/)|
|yi-34b|01-ai|1+ Providers|[huggingface.co](https://huggingface.co/01-ai/Yi-34B-Chat)|
@ -223,15 +227,16 @@ This document provides an overview of various AI providers and models, including
|tulu-3-1-8b|Ai2|1+ Providers|[allenai.org](https://allenai.org/documentation)|
|tulu-3-70b|Ai2|1+ Providers|[allenai.org](https://allenai.org/documentation)|
|olmoe-0125|Ai2|1+ Providers|[allenai.org](https://allenai.org/documentation)|
|lfm-40b|Liquid AI|1+ Providers|[liquid.ai](https://www.liquid.ai/liquid-foundation-models)|
|evil|Evil Mode - Experimental|1+ Providers|[]( )|
---
### Image Models
### Image generation models
| Model | Base Provider | Providers | Website |
|-------|---------------|-----------|---------|
|sdxl-turbo|Stability AI|2+ Providers|[huggingface.co](https://huggingface.co/stabilityai/sdxl-turbo)|
|sd-3.5|Stability AI|1+ Providers|[huggingface.co](https://huggingface.co/stabilityai/stable-diffusion-3.5-large)|
|flux|Black Forest Labs|3+ Providers|[github.com/black-forest-labs/flux](https://github.com/black-forest-labs/flux)|
|flux|Black Forest Labs|4+ Providers|[github.com/black-forest-labs/flux](https://github.com/black-forest-labs/flux)|
|flux-pro|Black Forest Labs|1+ Providers|[huggingface.co](https://huggingface.co/enhanceaiteam/FLUX.1-Pro)|
|flux-dev|Black Forest Labs|4+ Providers|[huggingface.co](https://huggingface.co/black-forest-labs/FLUX.1-dev)|
|flux-schnell|Black Forest Labs|4+ Providers|[huggingface.co](https://huggingface.co/black-forest-labs/FLUX.1-schnell)|
@ -239,7 +244,6 @@ This document provides an overview of various AI providers and models, including
|midjourney|Midjourney|1+ Providers|[docs.midjourney.com](https://docs.midjourney.com/docs/model-versions)|
## Conclusion and Usage Tips
This document provides a comprehensive overview of various AI providers and models available for text generation, image generation, and vision tasks. **When choosing a provider or model, consider the following factors:**
1. **Availability**: Check the status of the provider to ensure it's currently active and accessible.

9
etc/examples/vision_images.py
View file

@ -2,16 +2,13 @@ import g4f
import requests
from g4f.client import Client
from g4f.Provider.Blackbox import Blackbox
client = Client(
provider=Blackbox
)
client = Client()
# Processing remote image
remote_image = requests.get("https://raw.githubusercontent.com/xtekky/gpt4free/refs/heads/main/docs/images/cat.jpeg", stream=True).content
response_remote = client.chat.completions.create(
model=g4f.models.default,
model=g4f.models.default_vision,
messages=[
{"role": "user", "content": "What are on this image?"}
],
@ -25,7 +22,7 @@ print("\n" + "-"*50 + "\n") # Separator
# Processing local image
local_image = open("docs/images/cat.jpeg", "rb")
response_local = client.chat.completions.create(
model=g4f.models.default,
model=g4f.models.default_vision,
messages=[
{"role": "user", "content": "What are on this image?"}
],

3
g4f/Provider/AllenAI.py
View file

@ -48,6 +48,9 @@ class AllenAI(AsyncGeneratorProvider, ProviderModelMixin):
"olmo-2-13b": "OLMo-2-1124-13B-Instruct",
"tulu-3-1-8b": "tulu-3-1-8b",
"tulu-3-70b": "Llama-3-1-Tulu-3-70B",
"llama-3.1-405b": "Llama-3-1-Tulu-3-70B",
"llama-3.1-8b": "tulu-3-1-8b",
"llama-3.1-70b": "Llama-3-1-Tulu-3-70B",
}
@classmethod

524
g4f/Provider/Blackbox.py
View file

@ -1,6 +1,7 @@
from __future__ import annotations
from aiohttp import ClientSession
import os
import re
import json
import random
@ -8,6 +9,7 @@ import string
import base64
from pathlib import Path
from typing import Optional
from datetime import datetime, timedelta
from ..typing import AsyncResult, Messages, ImagesType
from ..requests.raise_for_status import raise_for_status
@ -17,6 +19,7 @@ from ..cookies import get_cookies_dir
from .helper import format_prompt, format_image_prompt
from ..providers.response import JsonConversation, ImageResponse
from ..errors import ModelNotSupportedError
from .. import debug
class Conversation(JsonConversation):
validated_value: str = None
@ -38,79 +41,157 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
default_model = "blackboxai"
default_vision_model = default_model
default_image_model = 'ImageGeneration'
default_image_model = 'flux'
# Completely free models
fallback_models = [
"blackboxai",
"gpt-4o-mini",
"GPT-4o",
"o1",
"o3-mini",
"Claude-sonnet-3.7",
"DeepSeek-V3",
"DeepSeek-R1",
"DeepSeek-LLM-Chat-(67B)",
# Image models
"flux",
# Trending agent modes
'Python Agent',
'HTML Agent',
'Builder Agent',
'Java Agent',
'JavaScript Agent',
'React Agent',
'Android Agent',
'Flutter Agent',
'Next.js Agent',
'AngularJS Agent',
'Swift Agent',
'MongoDB Agent',
'PyTorch Agent',
'Xcode Agent',
'Azure Agent',
'Bitbucket Agent',
'DigitalOcean Agent',
'Docker Agent',
'Electron Agent',
'Erlang Agent',
'FastAPI Agent',
'Firebase Agent',
'Flask Agent',
'Git Agent',
'Gitlab Agent',
'Go Agent',
'Godot Agent',
'Google Cloud Agent',
'Heroku Agent'
]
image_models = [default_image_model]
vision_models = [default_vision_model, 'gpt-4o', 'o1', 'o3-mini', 'gemini-pro', 'gemini-1.5-flash', 'llama-3.1-8b', 'llama-3.1-70b', 'llama-3.1-405b', 'gemini-2.0-flash', 'deepseek-v3']
vision_models = [default_vision_model, 'GPT-4o', 'o1', 'o3-mini', 'Gemini-PRO', 'Gemini Agent', 'llama-3.1-8b Agent', 'llama-3.1-70b Agent', 'llama-3.1-405 Agent', 'Gemini-Flash-2.0', 'DeepSeek-V3']
userSelectedModel = ['gpt-4o', 'o1', 'o3-mini', 'gemini-pro', 'claude-sonnet-3.7', 'deepseek-v3', 'deepseek-r1', 'blackboxai-pro', 'Meta-Llama-3.3-70B-Instruct-Turbo', 'Mistral-Small-24B-Instruct-2501', 'DeepSeek-LLM-Chat-(67B)', 'dbrx-instruct', 'Qwen-QwQ-32B-Preview', 'Nous-Hermes-2-Mixtral-8x7B-DPO', 'gemini-2.0-flash']
userSelectedModel = ['GPT-4o', 'o1', 'o3-mini', 'Gemini-PRO', 'Claude-sonnet-3.7', 'DeepSeek-V3', 'DeepSeek-R1', 'Meta-Llama-3.3-70B-Instruct-Turbo', 'Mistral-Small-24B-Instruct-2501', 'DeepSeek-LLM-Chat-(67B)', 'DBRX-Instruct', 'Qwen-QwQ-32B-Preview', 'Nous-Hermes-2-Mixtral-8x7B-DPO', 'Gemini-Flash-2.0']
# Agent mode configurations
agentMode = {
'deepseek-v3': {'mode': True, 'id': "deepseek-chat", 'name': "DeepSeek-V3"},
'deepseek-r1': {'mode': True, 'id': "deepseek-reasoner", 'name': "DeepSeek-R1"},
'GPT-4o': {'mode': True, 'id': "GPT-4o", 'name': "GPT-4o"},
'Gemini-PRO': {'mode': True, 'id': "Gemini-PRO", 'name': "Gemini-PRO"},
'Claude-sonnet-3.7': {'mode': True, 'id': "Claude-sonnet-3.7", 'name': "Claude-sonnet-3.7"},
'DeepSeek-V3': {'mode': True, 'id': "deepseek-chat", 'name': "DeepSeek-V3"},
'DeepSeek-R1': {'mode': True, 'id': "deepseek-reasoner", 'name': "DeepSeek-R1"},
'Meta-Llama-3.3-70B-Instruct-Turbo': {'mode': True, 'id': "meta-llama/Llama-3.3-70B-Instruct-Turbo", 'name': "Meta-Llama-3.3-70B-Instruct-Turbo"},
'Gemini-Flash-2.0': {'mode': True, 'id': "Gemini/Gemini-Flash-2.0", 'name': "Gemini-Flash-2.0"},
'Mistral-Small-24B-Instruct-2501': {'mode': True, 'id': "mistralai/Mistral-Small-24B-Instruct-2501", 'name': "Mistral-Small-24B-Instruct-2501"},
'DeepSeek-LLM-Chat-(67B)': {'mode': True, 'id': "deepseek-ai/deepseek-llm-67b-chat", 'name': "DeepSeek-LLM-Chat-(67B)"},
'dbrx-instruct': {'mode': True, 'id': "databricks/dbrx-instruct", 'name': "DBRX-Instruct"},
'DBRX-Instruct': {'mode': True, 'id': "databricks/dbrx-instruct", 'name': "DBRX-Instruct"},
'Qwen-QwQ-32B-Preview': {'mode': True, 'id': "Qwen/QwQ-32B-Preview", 'name': "Qwen-QwQ-32B-Preview"},
'Nous-Hermes-2-Mixtral-8x7B-DPO': {'mode': True, 'id': "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", 'name': "Nous-Hermes-2-Mixtral-8x7B-DPO"},
'gemini-2.0-flash': {'mode': True, 'id': "Gemini/Gemini-Flash-2.0", 'name': "Gemini-Flash-2.0"},
}
# Trending agent modes
trendingAgentMode = {
"gemini-1.5-flash": {'mode': True, 'id': 'Gemini'},
"llama-3.1-8b": {'mode': True, 'id': "llama-3.1-8b"},
'llama-3.1-70b': {'mode': True, 'id': "llama-3.1-70b"},
'llama-3.1-405b': {'mode': True, 'id': "llama-3.1-405"},
'Python Agent': {'mode': True, 'id': "Python Agent"},
'Java Agent': {'mode': True, 'id': "Java Agent"},
'JavaScript Agent': {'mode': True, 'id': "JavaScript Agent"},
'HTML Agent': {'mode': True, 'id': "HTML Agent"},
'Google Cloud Agent': {'mode': True, 'id': "Google Cloud Agent"},
'Android Developer': {'mode': True, 'id': "Android Developer"},
'Swift Developer': {'mode': True, 'id': "Swift Developer"},
'Next.js Agent': {'mode': True, 'id': "Next.js Agent"},
'MongoDB Agent': {'mode': True, 'id': "MongoDB Agent"},
'PyTorch Agent': {'mode': True, 'id': "PyTorch Agent"},
'React Agent': {'mode': True, 'id': "React Agent"},
'Xcode Agent': {'mode': True, 'id': "Xcode Agent"},
'blackboxai-pro': {'mode': True, 'id': "BLACKBOXAI-PRO"},
'Heroku Agent': {'mode': True, 'id': "Heroku Agent"},
'Godot Agent': {'mode': True, 'id': "Godot Agent"},
'Go Agent': {'mode': True, 'id': "Go Agent"},
'Gitlab Agent': {'mode': True, 'id': "Gitlab Agent"},
'Git Agent': {'mode': True, 'id': "Git Agent"},
'Flask Agent': {'mode': True, 'id': "Flask Agent"},
'Firebase Agent': {'mode': True, 'id': "Firebase Agent"},
'FastAPI Agent': {'mode': True, 'id': "FastAPI Agent"},
'Erlang Agent': {'mode': True, 'id': "Erlang Agent"},
'Electron Agent': {'mode': True, 'id': "Electron Agent"},
'Docker Agent': {'mode': True, 'id': "Docker Agent"},
'DigitalOcean Agent': {'mode': True, 'id': "DigitalOcean Agent"},
'Bitbucket Agent': {'mode': True, 'id': "Bitbucket Agent"},
'Azure Agent': {'mode': True, 'id': "Azure Agent"},
'Flutter Agent': {'mode': True, 'id': "Flutter Agent"},
'Youtube Agent': {'mode': True, 'id': "Youtube Agent"},
'builder Agent': {'mode': True, 'id': "builder Agent"},
"Gemini Agent": {'mode': True, 'id': 'gemini'},
"llama-3.1-405 Agent": {'mode': True, 'id': "llama-3.1-405"},
'llama-3.1-70b Agent': {'mode': True, 'id': "llama-3.1-70b"},
'llama-3.1-8b Agent': {'mode': True, 'id': "llama-3.1-8b"},
'Python Agent': {'mode': True, 'id': "python"},
'HTML Agent': {'mode': True, 'id': "html"},
'Builder Agent': {'mode': True, 'id': "builder"},
'Java Agent': {'mode': True, 'id': "java"},
'JavaScript Agent': {'mode': True, 'id': "javascript"},
'React Agent': {'mode': True, 'id': "react"},
'Android Agent': {'mode': True, 'id': "android"},
'Flutter Agent': {'mode': True, 'id': "flutter"},
'Next.js Agent': {'mode': True, 'id': "next.js"},
'AngularJS Agent': {'mode': True, 'id': "angularjs"},
'Swift Agent': {'mode': True, 'id': "swift"},
'MongoDB Agent': {'mode': True, 'id': "mongodb"},
'PyTorch Agent': {'mode': True, 'id': "pytorch"},
'Xcode Agent': {'mode': True, 'id': "xcode"},
'Azure Agent': {'mode': True, 'id': "azure"},
'Bitbucket Agent': {'mode': True, 'id': "bitbucket"},
'DigitalOcean Agent': {'mode': True, 'id': "digitalocean"},
'Docker Agent': {'mode': True, 'id': "docker"},
'Electron Agent': {'mode': True, 'id': "electron"},
'Erlang Agent': {'mode': True, 'id': "erlang"},
'FastAPI Agent': {'mode': True, 'id': "fastapi"},
'Firebase Agent': {'mode': True, 'id': "firebase"},
'Flask Agent': {'mode': True, 'id': "flask"},
'Git Agent': {'mode': True, 'id': "git"},
'Gitlab Agent': {'mode': True, 'id': "gitlab"},
'Go Agent': {'mode': True, 'id': "go"},
'Godot Agent': {'mode': True, 'id': "godot"},
'Google Cloud Agent': {'mode': True, 'id': "googlecloud"},
'Heroku Agent': {'mode': True, 'id': "heroku"},
}
models = list(dict.fromkeys([default_model, *userSelectedModel, *image_models, *list(agentMode.keys()), *list(trendingAgentMode.keys())]))
# Complete list of all models (for authorized users)
_all_models = list(dict.fromkeys([
default_model,
*userSelectedModel,
*image_models,
*list(agentMode.keys()),
*list(trendingAgentMode.keys())
]))
model_aliases = {
"gemini-1.5-flash": "gemini-1.5-flash",
"gemini-1.5-pro": "gemini-pro",
"llama-3.3-70b": "Meta-Llama-3.3-70B-Instruct-Turbo",
"mixtral-small-28b": "Mistral-Small-24B-Instruct-2501",
"deepseek-chat": "DeepSeek-LLM-Chat-(67B)",
"qwq-32b": "Qwen-QwQ-32B-Preview",
"hermes-2-dpo": "Nous-Hermes-2-Mixtral-8x7B-DPO",
"claude-3.7-sonnet": "claude-sonnet-3.7",
"flux": "ImageGeneration",
}
@classmethod
def generate_session(cls, id_length: int = 21, days_ahead: int = 365) -> dict:
"""
Generate a dynamic session with proper ID and expiry format.
ENCRYPTED_SESSION = "eyJ1c2VyIjogeyJuYW1lIjogIkJMQUNLQk9YIEFJIiwgImVtYWlsIjogImdpc2VsZUBibGFja2JveC5haSIsICJpbWFnZSI6ICJodHRwczovL3l0My5nb29nbGV1c2VyY29udGVudC5jb20vQjd6RVlVSzUxWnNQYmFSUFVhMF9ZbnQ1WV9URFZoTE4tVjAzdndRSHM0eF96a2g4a1psLXkxcXFxb3hoeFFzcS1wUVBHS0R0WFE9czE2MC1jLWstYzB4MDBmZmZmZmYtbm8tcmoifSwgImV4cGlyZXMiOiBudWxsfQ=="
ENCRYPTED_SUBSCRIPTION_CACHE = "eyJzdGF0dXMiOiAiUFJFTUlVTSIsICJleHBpcnlUaW1lc3RhbXAiOiBudWxsLCAibGFzdENoZWNrZWQiOiBudWxsLCAiaXNUcmlhbFN1YnNjcmlwdGlvbiI6IHRydWV9"
ENCRYPTED_IS_PREMIUM = "dHJ1ZQ=="
Args:
id_length: Length of the numeric ID (default: 21)
days_ahead: Number of days ahead for expiry (default: 365)
Returns:
dict: A session dictionary with user information and expiry
"""
# Generate numeric ID
numeric_id = ''.join(random.choice('0123456789') for _ in range(id_length))
# Generate future expiry date
future_date = datetime.now() + timedelta(days=days_ahead)
expiry = future_date.strftime('%Y-%m-%dT%H:%M:%S.%f')[:-3] + 'Z'
# Decode the encoded email
encoded_email = "Z2lzZWxlQGJsYWNrYm94LmFp" # Base64 encoded email
email = base64.b64decode(encoded_email).decode('utf-8')
# Generate random image ID for the new URL format
chars = string.ascii_letters + string.digits + "-"
random_img_id = ''.join(random.choice(chars) for _ in range(48))
image_url = f"https://lh3.googleusercontent.com/a/ACg8oc{random_img_id}=s96-c"
return {
"user": {
"name": "BLACKBOX AI",
"email": email,
"image": image_url,
"id": numeric_id
},
"expires": expiry
}
@classmethod
async def fetch_validated(cls, url: str = "https://www.blackbox.ai", force_refresh: bool = False) -> Optional[str]:
@ -123,7 +204,7 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
if data.get('validated_value'):
return data['validated_value']
except Exception as e:
print(f"Error reading cache: {e}")
debug.log(f"Blackbox: Error reading cache: {e}")
js_file_pattern = r'static/chunks/\d{4}-[a-fA-F0-9]+\.js'
uuid_pattern = r'["\']([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12})["\']'
@ -158,12 +239,12 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
with open(cache_file, 'w') as f:
json.dump({'validated_value': validated_value}, f)
except Exception as e:
print(f"Error writing cache: {e}")
debug.log(f"Blackbox: Error writing cache: {e}")
return validated_value
except Exception as e:
print(f"Error retrieving validated_value: {e}")
debug.log(f"Blackbox: Error retrieving validated_value: {e}")
return None
@ -172,19 +253,189 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
chars = string.ascii_letters + string.digits
return ''.join(random.choice(chars) for _ in range(length))
@staticmethod
def decrypt_data(encrypted_data):
try:
return json.loads(base64.b64decode(encrypted_data).decode('utf-8'))
except:
return None
@classmethod
def get_models(cls) -> list:
"""
Returns a list of available models based on authorization status.
Authorized users get the full list of models.
Unauthorized users only get fallback_models.
"""
# Check if there are valid session data in HAR files
has_premium_access = cls._check_premium_access()
@staticmethod
def decrypt_bool(encrypted_data):
if has_premium_access:
# For authorized users - all models
debug.log(f"Blackbox: Returning full model list with {len(cls._all_models)} models")
return cls._all_models
else:
# For demo accounts - only free models
debug.log(f"Blackbox: Returning free model list with {len(cls.fallback_models)} models")
return cls.fallback_models
@classmethod
def _check_premium_access(cls) -> bool:
"""
Checks for an authorized session in HAR files.
Returns True if a valid session is found that differs from the demo.
"""
try:
return base64.b64decode(encrypted_data).decode('utf-8').lower() == 'true'
except:
har_dir = get_cookies_dir()
if not os.access(har_dir, os.R_OK):
return False
for root, _, files in os.walk(har_dir):
for file in files:
if file.endswith(".har"):
try:
with open(os.path.join(root, file), 'rb') as f:
har_data = json.load(f)
for entry in har_data['log']['entries']:
# Only check requests to blackbox API
if 'blackbox.ai/api' in entry['request']['url']:
if 'response' in entry and 'content' in entry['response']:
content = entry['response']['content']
if ('text' in content and
isinstance(content['text'], str) and
'"user"' in content['text'] and
'"email"' in content['text']):
try:
# Process request text
text = content['text'].strip()
if text.startswith('{') and text.endswith('}'):
text = text.replace('\\"', '"')
session_data = json.loads(text)
# Check if this is a valid session
if (isinstance(session_data, dict) and
'user' in session_data and
'email' in session_data['user']):
# Check if this is not a demo session
demo_session = cls.generate_session()
if (session_data['user'].get('email') !=
demo_session['user'].get('email')):
# This is not a demo session, so user has premium access
return True
except:
pass
except:
pass
return False
except Exception as e:
debug.log(f"Blackbox: Error checking premium access: {e}")
return False
# Initialize models with fallback_models
models = fallback_models
model_aliases = {
"gpt-4o": "GPT-4o",
"claude-3.7-sonnet": "Claude-sonnet-3.7",
"deepseek-v3": "DeepSeek-V3",
"deepseek-r1": "DeepSeek-R1",
"deepseek-chat": "DeepSeek-LLM-Chat-(67B)",
}
@classmethod
def generate_session(cls, id_length: int = 21, days_ahead: int = 365) -> dict:
"""
Generate a dynamic session with proper ID and expiry format.
Args:
id_length: Length of the numeric ID (default: 21)
days_ahead: Number of days ahead for expiry (default: 365)
Returns:
dict: A session dictionary with user information and expiry
"""
# Generate numeric ID
numeric_id = ''.join(random.choice('0123456789') for _ in range(id_length))
# Generate future expiry date
future_date = datetime.now() + timedelta(days=days_ahead)
expiry = future_date.strftime('%Y-%m-%dT%H:%M:%S.%f')[:-3] + 'Z'
# Decode the encoded email
encoded_email = "Z2lzZWxlQGJsYWNrYm94LmFp" # Base64 encoded email
email = base64.b64decode(encoded_email).decode('utf-8')
# Generate random image ID for the new URL format
chars = string.ascii_letters + string.digits + "-"
random_img_id = ''.join(random.choice(chars) for _ in range(48))
image_url = f"https://lh3.googleusercontent.com/a/ACg8oc{random_img_id}=s96-c"
return {
"user": {
"name": "BLACKBOX AI",
"email": email,
"image": image_url,
"id": numeric_id
},
"expires": expiry
}
@classmethod
async def fetch_validated(cls, url: str = "https://www.blackbox.ai", force_refresh: bool = False) -> Optional[str]:
cache_file = Path(get_cookies_dir()) / 'blackbox.json'
if not force_refresh and cache_file.exists():
try:
with open(cache_file, 'r') as f:
data = json.load(f)
if data.get('validated_value'):
return data['validated_value']
except Exception as e:
debug.log(f"Blackbox: Error reading cache: {e}")
js_file_pattern = r'static/chunks/\d{4}-[a-fA-F0-9]+\.js'
uuid_pattern = r'["\']([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12})["\']'
def is_valid_context(text: str) -> bool:
return any(char + '=' in text for char in 'abcdefghijklmnopqrstuvwxyz')
async with ClientSession() as session:
try:
async with session.get(url) as response:
if response.status != 200:
return None
page_content = await response.text()
js_files = re.findall(js_file_pattern, page_content)
for js_file in js_files:
js_url = f"{url}/_next/{js_file}"
async with session.get(js_url) as js_response:
if js_response.status == 200:
js_content = await js_response.text()
for match in re.finditer(uuid_pattern, js_content):
start = max(0, match.start() - 10)
end = min(len(js_content), match.end() + 10)
context = js_content[start:end]
if is_valid_context(context):
validated_value = match.group(1)
cache_file.parent.mkdir(exist_ok=True)
try:
with open(cache_file, 'w') as f:
json.dump({'validated_value': validated_value}, f)
except Exception as e:
debug.log(f"Blackbox: Error writing cache: {e}")
return validated_value
except Exception as e:
debug.log(f"Blackbox: Error retrieving validated_value: {e}")
return None
@classmethod
def generate_id(cls, length: int = 7) -> str:
chars = string.ascii_letters + string.digits
return ''.join(random.choice(chars) for _ in range(length))
@classmethod
async def create_async_generator(
@ -212,30 +463,6 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
}
async with ClientSession(headers=headers) as session:
if model in "ImageGeneration":
prompt = format_image_prompt(messages, prompt)
data = {
"query": format_image_prompt(messages, prompt),
"agentMode": True
}
headers['content-type'] = 'text/plain;charset=UTF-8'
async with session.post(
"https://www.blackbox.ai/api/image-generator",
json=data,
proxy=proxy,
headers=headers
) as response:
await raise_for_status(response)
response_json = await response.json()
if "markdown" in response_json:
image_url_match = re.search(r'!\[.*?\]\((.*?)\)', response_json["markdown"])
if image_url_match:
image_url = image_url_match.group(1)
yield ImageResponse(images=[image_url], alt=format_image_prompt(messages, prompt))
return
if conversation is None or not hasattr(conversation, "chat_id"):
conversation = Conversation(model)
conversation.validated_value = await cls.fetch_validated()
@ -265,6 +492,71 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
"title": ""
}
# Try to get session data from HAR files
session_data = cls.generate_session() # Default fallback
session_found = False
# Look for HAR session data
har_dir = get_cookies_dir()
if os.access(har_dir, os.R_OK):
for root, _, files in os.walk(har_dir):
for file in files:
if file.endswith(".har"):
try:
with open(os.path.join(root, file), 'rb') as f:
har_data = json.load(f)
for entry in har_data['log']['entries']:
# Only look at blackbox API responses
if 'blackbox.ai/api' in entry['request']['url']:
# Look for a response that has the right structure
if 'response' in entry and 'content' in entry['response']:
content = entry['response']['content']
# Look for both regular and Google auth session formats
if ('text' in content and
isinstance(content['text'], str) and
'"user"' in content['text'] and
'"email"' in content['text'] and
'"expires"' in content['text']):
try:
# Remove any HTML or other non-JSON content
text = content['text'].strip()
if text.startswith('{') and text.endswith('}'):
# Replace escaped quotes
text = text.replace('\\"', '"')
har_session = json.loads(text)
# Check if this is a valid session object (supports both regular and Google auth)
if (isinstance(har_session, dict) and
'user' in har_session and
'email' in har_session['user'] and
'expires' in har_session):
file_path = os.path.join(root, file)
debug.log(f"Blackbox: Found session in HAR file")
session_data = har_session
session_found = True
break
except json.JSONDecodeError as e:
# Only print error for entries that truly look like session data
if ('"user"' in content['text'] and
'"email"' in content['text']):
debug.log(f"Blackbox: Error parsing likely session data: {e}")
if session_found:
break
except Exception as e:
debug.log(f"Blackbox: Error reading HAR file: {e}")
if session_found:
break
if session_found:
break
data = {
"messages": current_messages,
"agentMode": cls.agentMode.get(model, {}) if model in cls.agentMode else {},
@ -288,7 +580,7 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
"mobileClient": False,
"userSelectedModel": model if model in cls.userSelectedModel else None,
"validated": conversation.validated_value,
"imageGenerationMode": False,
"imageGenerationMode": model == cls.default_image_model,
"webSearchModePrompt": False,
"deepSearchMode": False,
"domains": None,
@ -301,23 +593,57 @@ class Blackbox(AsyncGeneratorProvider, ProviderModelMixin):
"additionalInfo": "",
"enableNewChats": False
},
"session": cls.decrypt_data(cls.ENCRYPTED_SESSION),
"isPremium": cls.decrypt_bool(cls.ENCRYPTED_IS_PREMIUM),
"subscriptionCache": cls.decrypt_data(cls.ENCRYPTED_SUBSCRIPTION_CACHE),
"session": session_data if session_data else cls.generate_session(),
"isPremium": True,
"subscriptionCache": None,
"beastMode": False,
"webSearchMode": False
}
# Add debugging before making the API call
if isinstance(session_data, dict) and 'user' in session_data:
# Генеруємо демо-сесію для порівняння
demo_session = cls.generate_session()
is_demo = False
if demo_session and isinstance(demo_session, dict) and 'user' in demo_session:
if session_data['user'].get('email') == demo_session['user'].get('email'):
is_demo = True
if is_demo:
debug.log(f"Blackbox: Making API request with built-in Developer Premium Account")
else:
user_email = session_data['user'].get('email', 'unknown')
debug.log(f"Blackbox: Making API request with HAR session email: {user_email}")
# Continue with the API request and async generator behavior
async with session.post(cls.api_endpoint, json=data, proxy=proxy) as response:
await raise_for_status(response)
# Collect the full response
full_response = []
async for chunk in response.content.iter_any():
if chunk:
chunk_text = chunk.decode()
full_response.append(chunk_text)
yield chunk_text
# Only yield chunks for non-image models
if model != cls.default_image_model:
yield chunk_text
full_response_text = ''.join(full_response)
# For image models, check for image markdown
if model == cls.default_image_model:
image_url_match = re.search(r'!\[.*?\]\((.*?)\)', full_response_text)
if image_url_match:
image_url = image_url_match.group(1)
yield ImageResponse(images=[image_url], alt=format_image_prompt(messages, prompt))
return
# Handle conversation history once, in one place
if return_conversation:
full_response_text = ''.join(full_response)
conversation.message_history.append({"role": "assistant", "content": full_response_text})
yield conversation
# For image models that didn't produce an image, fall back to text response
elif model == cls.default_image_model:
yield full_response_text

2
g4f/Provider/Copilot.py
View file

@ -48,7 +48,7 @@ class Copilot(AbstractProvider, ProviderModelMixin):
models = [default_model]
model_aliases = {
"gpt-4": default_model,
"gpt-4o": default_model,
"o1": default_model,
}
websocket_url = "wss://copilot.microsoft.com/c/api/chat?api-version=2"

2
g4f/Provider/DeepInfraChat.py
View file

@ -41,7 +41,7 @@ class DeepInfraChat(OpenaiTemplate):
"llama-3.3-70b": "meta-llama/Llama-3.3-70B-Instruct-Turbo",
"llama-3.3-70b": "meta-llama/Llama-3.3-70B-Instruct",
"deepseek-v3": default_model,
"mixtral-small-28b": "mistralai/Mistral-Small-24B-Instruct-2501",
"mixtral-small-24b": "mistralai/Mistral-Small-24B-Instruct-2501",
"deepseek-r1": "deepseek-ai/DeepSeek-R1-Turbo",
"deepseek-r1": "deepseek-ai/DeepSeek-R1",
"deepseek-r1-distill-llama": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",

67
g4f/Provider/Dynaspark.py Normal file
View file

@ -0,0 +1,67 @@
from __future__ import annotations
import json
from aiohttp import ClientSession, FormData
from ..typing import AsyncResult, Messages, ImagesType
from .base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..requests.raise_for_status import raise_for_status
from ..image import to_data_uri, to_bytes, is_accepted_format
from .helper import format_prompt
class Dynaspark(AsyncGeneratorProvider, ProviderModelMixin):
url = "https://dynaspark.onrender.com"
login_url = None
api_endpoint = "https://dynaspark.onrender.com/generate_response"
working = True
needs_auth = False
use_nodriver = True
supports_stream = True
supports_system_message = False
supports_message_history = False
default_model = 'gemini-1.5-flash'
default_vision_model = default_model
vision_models = [default_vision_model, 'gemini-1.5-flash-8b', 'gemini-2.0-flash', 'gemini-2.0-flash-lite']
models = vision_models
model_aliases = {
"gemini-1.5-flash": "gemini-1.5-flash-8b",
"gemini-2.0-flash": "gemini-2.0-flash-lite",
}
@classmethod
async def create_async_generator(
cls,
model: str,
messages: Messages,
proxy: str = None,
images: ImagesType = None,
**kwargs
) -> AsyncResult:
headers = {
'accept': '*/*',
'accept-language': 'en-US,en;q=0.9',
'origin': 'https://dynaspark.onrender.com',
'referer': 'https://dynaspark.onrender.com/',
'user-agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36',
'x-requested-with': 'XMLHttpRequest'
}
async with ClientSession(headers=headers) as session:
form = FormData()
form.add_field('user_input', format_prompt(messages))
form.add_field('ai_model', model)
if images is not None and len(images) > 0:
image, image_name = images[0]
image_bytes = to_bytes(image)
form.add_field('file', image_bytes, filename=image_name, content_type=is_accepted_format(image_bytes))
async with session.post(f"{cls.api_endpoint}", data=form, proxy=proxy) as response:
await raise_for_status(response)
response_text = await response.text()
response_json = json.loads(response_text)
yield response_json["response"]

49
g4f/Provider/Goabror.py Normal file
View file

@ -0,0 +1,49 @@
from __future__ import annotations
import json
from aiohttp import ClientSession
from ..typing import AsyncResult, Messages
from .base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..requests.raise_for_status import raise_for_status
from .helper import format_prompt
class Goabror(AsyncGeneratorProvider, ProviderModelMixin):
url = "https://goabror.uz"
api_endpoint = "https://goabror.uz/api/gpt.php"
working = True
default_model = 'gpt-4'
models = [default_model]
@classmethod
async def create_async_generator(
cls,
model: str,
messages: Messages,
proxy: str = None,
**kwargs
) -> AsyncResult:
headers = {
'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.7',
'accept-language': 'en-US,en;q=0.9',
'user-agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36'
}
async with ClientSession(headers=headers) as session:
params = {
"user": format_prompt(messages)
}
async with session.get(f"{cls.api_endpoint}", params=params, proxy=proxy) as response:
await raise_for_status(response)
text_response = await response.text()
try:
json_response = json.loads(text_response)
if "data" in json_response:
yield json_response["data"]
else:
yield text_response
except json.JSONDecodeError:
yield text_response

180
g4f/Provider/LambdaChat.py Normal file
View file

@ -0,0 +1,180 @@
from __future__ import annotations
import json
import re
import uuid
from aiohttp import ClientSession, FormData
from ..typing import AsyncResult, Messages
from ..requests import raise_for_status
from .base_provider import AsyncGeneratorProvider, ProviderModelMixin
from .helper import format_prompt, get_last_user_message
from ..providers.response import JsonConversation, TitleGeneration, Reasoning
class LambdaChat(AsyncGeneratorProvider, ProviderModelMixin):
label = "Lambda Chat"
url = "https://lambda.chat"
conversation_url = f"{url}/conversation"
working = True
default_model = "deepseek-llama3.3-70b"
reasoning_model = "deepseek-r1"
models = [
default_model,
reasoning_model,
"hermes-3-llama-3.1-405b-fp8",
"hermes3-405b-fp8-128k",
"llama3.1-nemotron-70b-instruct",
"lfm-40b",
"llama3.3-70b-instruct-fp8"
]
model_aliases = {
"deepseek-v3": default_model,
"hermes-3": "hermes-3-llama-3.1-405b-fp8",
"hermes-3": "hermes3-405b-fp8-128k",
"nemotron-70b": "llama3.1-nemotron-70b-instruct",
"llama-3.3-70b": "llama3.3-70b-instruct-fp8"
}
@classmethod
async def create_async_generator(
cls, model: str, messages: Messages,
api_key: str = None,
proxy: str = None,
cookies: dict = None,
**kwargs
) -> AsyncResult:
model = cls.get_model(model)
headers = {
"Origin": cls.url,
"User-Agent": "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36",
"Accept": "*/*",
"Accept-Language": "en-US,en;q=0.9",
"Referer": cls.url,
"Sec-Fetch-Dest": "empty",
"Sec-Fetch-Mode": "cors",
"Sec-Fetch-Site": "same-origin",
"Priority": "u=1, i",
"Pragma": "no-cache",
"Cache-Control": "no-cache"
}
# Initialize cookies if not provided
if cookies is None:
cookies = {
"hf-chat": str(uuid.uuid4()) # Generate a session ID
}
async with ClientSession(headers=headers, cookies=cookies) as session:
# Step 1: Create a new conversation
data = {"model": model}
async with session.post(cls.conversation_url, json=data, proxy=proxy) as response:
await raise_for_status(response)
conversation_response = await response.json()
conversation_id = conversation_response["conversationId"]
# Update cookies with any new ones from the response
for cookie_name, cookie in response.cookies.items():
cookies[cookie_name] = cookie.value
# Step 2: Get data for this conversation to extract message ID
async with session.get(
f"{cls.conversation_url}/{conversation_id}/__data.json?x-sveltekit-invalidated=11",
proxy=proxy
) as response:
await raise_for_status(response)
response_text = await response.text()
# Update cookies again
for cookie_name, cookie in response.cookies.items():
cookies[cookie_name] = cookie.value
try:
data_line = response_text.splitlines()[0]
data_json = json.loads(data_line)
# Navigate to the data section containing message info
message_id = None
# For debugging, print the JSON structure
if "nodes" in data_json and len(data_json["nodes"]) > 1:
node = data_json["nodes"][1]
if "data" in node:
data = node["data"]
# Try to find the system message ID
if len(data) > 1 and isinstance(data[1], list) and len(data[1]) > 2:
for item in data[1]:
if isinstance(item, dict) and "id" in item:
# Found a potential message ID
message_id = item["id"]
break
if not message_id:
# Fallback: Just search for any UUID in the response
uuid_pattern = r"[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}"
uuids = re.findall(uuid_pattern, response_text)
if uuids:
message_id = uuids[0]
if not message_id:
raise ValueError("Could not find message ID in response")
except (IndexError, KeyError, ValueError, json.JSONDecodeError) as e:
raise RuntimeError(f"Failed to parse conversation data: {str(e)}")
# Step 3: Send the user message
user_message = get_last_user_message(messages)
# Prepare form data exactly
form_data = FormData()
form_data.add_field(
"data",
json.dumps({
"inputs": user_message,
"id": message_id,
"is_retry": False,
"is_continue": False,
"web_search": False,
"tools": []
}),
content_type="application/json"
)
async with session.post(
f"{cls.conversation_url}/{conversation_id}",
data=form_data,
proxy=proxy
) as response:
await raise_for_status(response)
async for chunk in response.content:
if not chunk:
continue
chunk_str = chunk.decode('utf-8', errors='ignore')
try:
data = json.loads(chunk_str)
except json.JSONDecodeError:
continue
# Handling different types of responses
if data.get("type") == "stream" and "token" in data:
token = data["token"].replace("\u0000", "")
if token:
yield token
elif data.get("type") == "title":
yield TitleGeneration(data.get("title", ""))
elif data.get("type") == "reasoning" and model == cls.reasoning_model: # Only process reasoning for reasoning_model
subtype = data.get("subtype")
token = data.get("token", "").replace("\u0000", "")
status = data.get("status", "")
if subtype == "stream" and token:
yield Reasoning(token=token)
elif subtype == "status" and status:
yield Reasoning(status=status)
elif data.get("type") == "finalAnswer":
break

9
g4f/Provider/PollinationsAI.py
View file

@ -21,13 +21,6 @@ DEFAULT_HEADERS = {
"accept": "*/*",
'accept-language': 'en-US,en;q=0.9',
"user-agent": "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36",
"priority": "u=1, i",
"sec-ch-ua": "\"Not(A:Brand\";v=\"99\", \"Google Chrome\";v=\"133\", \"Chromium\";v=\"133\"",
"sec-ch-ua-mobile": "?0",
"sec-ch-ua-platform": "\"Linux\"",
"sec-fetch-dest": "empty",
"sec-fetch-mode": "cors",
"sec-fetch-site": "same-site",
"referer": "https://pollinations.ai/",
"origin": "https://pollinations.ai",
}
@ -71,6 +64,8 @@ class PollinationsAI(AsyncGeneratorProvider, ProviderModelMixin):
"gemini-2.0": "gemini",
"gemini-2.0-flash": "gemini",
"gemini-2.0-flash-thinking": "gemini-thinking",
"deepseek-r1": "deepseek-r1-llama",
"gpt-4o-audio": "openai-audio",
### Image Models ###
"sdxl-turbo": "turbo",

184
g4f/Provider/Websim.py Normal file
View file

@ -0,0 +1,184 @@
from __future__ import annotations
import json
import random
import string
import asyncio
from aiohttp import ClientSession
from ..typing import AsyncResult, Messages
from .base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..requests.raise_for_status import raise_for_status
from ..errors import ResponseStatusError
from ..providers.response import ImageResponse
from .helper import format_prompt, format_image_prompt
class Websim(AsyncGeneratorProvider, ProviderModelMixin):
url = "https://websim.ai"
login_url = None
chat_api_endpoint = "https://websim.ai/api/v1/inference/run_chat_completion"
image_api_endpoint = "https://websim.ai/api/v1/inference/run_image_generation"
working = True
needs_auth = False
use_nodriver = False
supports_stream = False
supports_system_message = True
supports_message_history = True
default_model = 'gemini-1.5-pro'
default_image_model = 'flux'
image_models = [default_image_model]
models = [default_model, 'gemini-1.5-flash'] + image_models
@staticmethod
def generate_project_id(for_image=False):
"""
Generate a project ID in the appropriate format
For chat: format like 'ke3_xh5gai3gjkmruomu'
For image: format like 'kx0m131_rzz66qb2xoy7'
"""
chars = string.ascii_lowercase + string.digits
if for_image:
first_part = ''.join(random.choices(chars, k=7))
second_part = ''.join(random.choices(chars, k=12))
return f"{first_part}_{second_part}"
else:
prefix = ''.join(random.choices(chars, k=3))
suffix = ''.join(random.choices(chars, k=15))
return f"{prefix}_{suffix}"
@classmethod
async def create_async_generator(
cls,
model: str,
messages: Messages,
prompt: str = None,
proxy: str = None,
aspect_ratio: str = "1:1",
project_id: str = None,
**kwargs
) -> AsyncResult:
is_image_request = model in cls.image_models
if project_id is None:
project_id = cls.generate_project_id(for_image=is_image_request)
headers = {
'accept': '*/*',
'accept-language': 'en-US,en;q=0.9',
'content-type': 'text/plain;charset=UTF-8',
'origin': 'https://websim.ai',
'user-agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36',
'websim-flags;': ''
}
if is_image_request:
headers['referer'] = 'https://websim.ai/@ISWEARIAMNOTADDICTEDTOPILLOW/ai-image-prompt-generator'
async for result in cls._handle_image_request(
project_id=project_id,
messages=messages,
prompt=prompt,
aspect_ratio=aspect_ratio,
headers=headers,
proxy=proxy,
**kwargs
):
yield result
else:
headers['referer'] = 'https://websim.ai/@ISWEARIAMNOTADDICTEDTOPILLOW/zelos-ai-assistant'
async for result in cls._handle_chat_request(
project_id=project_id,
messages=messages,
headers=headers,
proxy=proxy,
**kwargs
):
yield result
@classmethod
async def _handle_image_request(
cls,
project_id: str,
messages: Messages,
prompt: str,
aspect_ratio: str,
headers: dict,
proxy: str = None,
**kwargs
) -> AsyncResult:
used_prompt = format_image_prompt(messages, prompt)
async with ClientSession(headers=headers) as session:
data = {
"project_id": project_id,
"prompt": used_prompt,
"aspect_ratio": aspect_ratio
}
async with session.post(f"{cls.image_api_endpoint}", json=data, proxy=proxy) as response:
await raise_for_status(response)
response_text = await response.text()
response_json = json.loads(response_text)
image_url = response_json.get("url")
if image_url:
yield ImageResponse(images=[image_url], alt=used_prompt)
@classmethod
async def _handle_chat_request(
cls,
project_id: str,
messages: Messages,
headers: dict,
proxy: str = None,
**kwargs
) -> AsyncResult:
max_retries = 3
retry_count = 0
last_error = None
while retry_count < max_retries:
try:
async with ClientSession(headers=headers) as session:
data = {
"project_id": project_id,
"messages": messages
}
async with session.post(f"{cls.chat_api_endpoint}", json=data, proxy=proxy) as response:
if response.status == 429:
response_text = await response.text()
last_error = ResponseStatusError(f"Response {response.status}: {response_text}")
retry_count += 1
if retry_count < max_retries:
wait_time = 2 ** retry_count
await asyncio.sleep(wait_time)
continue
else:
raise last_error
await raise_for_status(response)
response_text = await response.text()
try:
response_json = json.loads(response_text)
content = response_json.get("content", "")
yield content.strip()
break
except json.JSONDecodeError:
yield response_text
break
except ResponseStatusError as e:
if "Rate limit exceeded" in str(e) and retry_count < max_retries:
retry_count += 1
wait_time = 2 ** retry_count
await asyncio.sleep(wait_time)
else:
if retry_count >= max_retries:
raise e
else:
raise
except Exception as e:
raise

5
g4f/Provider/__init__.py
View file

@ -24,12 +24,15 @@ from .Cloudflare import Cloudflare
from .Copilot import Copilot
from .DDG import DDG
from .DeepInfraChat import DeepInfraChat
from .Dynaspark import Dynaspark
from .Free2GPT import Free2GPT
from .FreeGpt import FreeGpt
from .GizAI import GizAI
from .Glider import Glider
from .Goabror import Goabror
from .ImageLabs import ImageLabs
from .Jmuz import Jmuz
from .LambdaChat import LambdaChat
from .Liaobots import Liaobots
from .OIVSCode import OIVSCode
from .PerplexityLabs import PerplexityLabs
@ -39,8 +42,10 @@ from .PollinationsAI import PollinationsAI
from .PollinationsImage import PollinationsImage
from .TeachAnything import TeachAnything
from .You import You
from .Websim import Websim
from .Yqcloud import Yqcloud
import sys
__modules__: list = [

5
g4f/Provider/hf_space/BlackForestLabsFlux1Dev.py → g4f/Provider/hf_space/BlackForestLabs_Flux1Dev.py
View file

@ -9,10 +9,11 @@ from ...requests import StreamSession
from ...errors import ResponseError
from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..helper import format_image_prompt
from .Janus_Pro_7B import get_zerogpu_token
from .DeepseekAI_JanusPro7b import get_zerogpu_token
from .raise_for_status import raise_for_status
class BlackForestLabsFlux1Dev(AsyncGeneratorProvider, ProviderModelMixin):
class BlackForestLabs_Flux1Dev(AsyncGeneratorProvider, ProviderModelMixin):
label = "BlackForestLabs Flux-1-Dev"
url = "https://black-forest-labs-flux-1-dev.hf.space"
space = "black-forest-labs/FLUX.1-dev"
referer = f"{url}/?__theme=light"

3
g4f/Provider/hf_space/BlackForestLabsFlux1Schnell.py → g4f/Provider/hf_space/BlackForestLabs_Flux1Schnell.py
View file

@ -10,7 +10,8 @@ from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..helper import format_image_prompt
from .raise_for_status import raise_for_status
class BlackForestLabsFlux1Schnell(AsyncGeneratorProvider, ProviderModelMixin):
class BlackForestLabs_Flux1Schnell(AsyncGeneratorProvider, ProviderModelMixin):
label = "BlackForestLabs Flux-1-Schnell"
url = "https://black-forest-labs-flux-1-schnell.hf.space"
api_endpoint = "https://black-forest-labs-flux-1-schnell.hf.space/call/infer"

8
g4f/Provider/hf_space/CohereForAI.py → g4f/Provider/hf_space/CohereForAI_C4AI_Command.py
View file

@ -9,21 +9,25 @@ from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..helper import format_prompt, get_last_user_message
from ...providers.response import JsonConversation, TitleGeneration
class CohereForAI(AsyncGeneratorProvider, ProviderModelMixin):
class CohereForAI_C4AI_Command(AsyncGeneratorProvider, ProviderModelMixin):
label = "CohereForAI C4AI Command"
url = "https://cohereforai-c4ai-command.hf.space"
conversation_url = f"{url}/conversation"
working = True
default_model = "command-r-plus-08-2024"
default_model = "command-a-03-2025"
models = [
default_model,
"command-r-plus-08-2024",
"command-r-08-2024",
"command-r-plus",
"command-r",
"command-r7b-12-2024",
"command-r7b-arabic-02-2025",
]
model_aliases = {
"command-a": "command-a-03-2025",
"command-r-plus": "command-r-plus-08-2024",
"command-r": "command-r-08-2024",
"command-r7b": "command-r7b-12-2024",

3
g4f/Provider/hf_space/Janus_Pro_7B.py → g4f/Provider/hf_space/DeepseekAI_JanusPro7b.py
View file

@ -19,7 +19,8 @@ from ...errors import ResponseError
from ... import debug
from .raise_for_status import raise_for_status
class Janus_Pro_7B(AsyncGeneratorProvider, ProviderModelMixin):
class DeepseekAI_JanusPro7b(AsyncGeneratorProvider, ProviderModelMixin):
label = "DeepseekAI Janus-Pro-7B"
space = "deepseek-ai/Janus-Pro-7B"
url = f"https://huggingface.co/spaces/{space}"
api_url = "https://deepseek-ai-janus-pro-7b.hf.space"

12
g4f/Provider/hf_space/G4F.py
View file

@ -8,16 +8,16 @@ import asyncio
from ...typing import AsyncResult, Messages
from ...providers.response import ImageResponse, Reasoning, JsonConversation
from ..helper import format_image_prompt, get_random_string
from .Janus_Pro_7B import Janus_Pro_7B, get_zerogpu_token
from .BlackForestLabsFlux1Dev import BlackForestLabsFlux1Dev
from .DeepseekAI_JanusPro7b import DeepseekAI_JanusPro7b, get_zerogpu_token
from .BlackForestLabs_Flux1Dev import BlackForestLabs_Flux1Dev
from .raise_for_status import raise_for_status
class FluxDev(BlackForestLabsFlux1Dev):
class FluxDev(BlackForestLabs_Flux1Dev):
url = "https://roxky-flux-1-dev.hf.space"
space = "roxky/FLUX.1-dev"
referer = f"{url}/?__theme=light"
class G4F(Janus_Pro_7B):
class G4F(DeepseekAI_JanusPro7b):
label = "G4F framework"
space = "roxky/Janus-Pro-7B"
url = f"https://huggingface.co/spaces/roxky/g4f-space"
@ -27,8 +27,8 @@ class G4F(Janus_Pro_7B):
default_model = "flux"
model_aliases = {"flux-schnell": default_model}
image_models = [Janus_Pro_7B.default_image_model, default_model, "flux-dev", *model_aliases.keys()]
models = [Janus_Pro_7B.default_model, *image_models]
image_models = [DeepseekAI_JanusPro7b.default_image_model, default_model, "flux-dev", *model_aliases.keys()]
models = [DeepseekAI_JanusPro7b.default_model, *image_models]
@classmethod
async def create_async_generator(

5
g4f/Provider/hf_space/Phi_4.py → g4f/Provider/hf_space/Microsoft_Phi_4.py
View file

@ -12,10 +12,11 @@ from ...requests.raise_for_status import raise_for_status
from ...image import to_bytes, is_accepted_format, is_data_an_wav
from ...errors import ResponseError
from ... import debug
from .Janus_Pro_7B import get_zerogpu_token
from .DeepseekAI_JanusPro7b import get_zerogpu_token
from .raise_for_status import raise_for_status
class Phi_4(AsyncGeneratorProvider, ProviderModelMixin):
class Microsoft_Phi_4(AsyncGeneratorProvider, ProviderModelMixin):
label = "Microsoft Phi-4"
space = "microsoft/phi-4-multimodal"
url = f"https://huggingface.co/spaces/{space}"
api_url = "https://microsoft-phi-4-multimodal.hf.space"

1
g4f/Provider/hf_space/Qwen_QVQ_72B.py
View file

@ -11,6 +11,7 @@ from ..helper import format_prompt, get_random_string
from ...image import to_bytes, is_accepted_format
class Qwen_QVQ_72B(AsyncGeneratorProvider, ProviderModelMixin):
label = "Qwen QVQ-72B"
url = "https://qwen-qvq-72b-preview.hf.space"
api_endpoint = "/gradio_api/call/generate"

1
g4f/Provider/hf_space/Qwen_Qwen_2_5M_Demo.py
View file

@ -12,6 +12,7 @@ from ..helper import get_last_user_message
from ... import debug
class Qwen_Qwen_2_5M_Demo(AsyncGeneratorProvider, ProviderModelMixin):
label = "Qwen Qwen-2.5M-Demo"
url = "https://qwen-qwen2-5-1m-demo.hf.space"
api_endpoint = f"{url}/run/predict?__theme=light"

1
g4f/Provider/hf_space/Qwen_Qwen_2_72B_Instruct.py
View file

@ -11,6 +11,7 @@ from ..helper import format_prompt
from ... import debug
class Qwen_Qwen_2_72B_Instruct(AsyncGeneratorProvider, ProviderModelMixin):
label = "Qwen Qwen-2.72B-Instruct"
url = "https://qwen-qwen2-72b-instruct.hf.space"
api_endpoint = "https://qwen-qwen2-72b-instruct.hf.space/queue/join?"

3
g4f/Provider/hf_space/StableDiffusion35Large.py → g4f/Provider/hf_space/StabilityAI_SD35Large.py
View file

@ -9,7 +9,8 @@ from ...errors import ResponseError
from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..helper import format_image_prompt
class StableDiffusion35Large(AsyncGeneratorProvider, ProviderModelMixin):
class StabilityAI_SD35Large(AsyncGeneratorProvider, ProviderModelMixin):
label = "StabilityAI SD-3.5-Large"
url = "https://stabilityai-stable-diffusion-3-5-large.hf.space"
api_endpoint = "/gradio_api/call/infer"

3
g4f/Provider/hf_space/VoodoohopFlux1Schnell.py → g4f/Provider/hf_space/Voodoohop_Flux1Schnell.py
View file

@ -10,7 +10,8 @@ from ...requests.raise_for_status import raise_for_status
from ..helper import format_image_prompt
from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
class VoodoohopFlux1Schnell(AsyncGeneratorProvider, ProviderModelMixin):
class Voodoohop_Flux1Schnell(AsyncGeneratorProvider, ProviderModelMixin):
label = "Voodoohop Flux-1-Schnell"
url = "https://voodoohop-flux-1-schnell.hf.space"
api_endpoint = "https://voodoohop-flux-1-schnell.hf.space/call/infer"

35
g4f/Provider/hf_space/__init__.py
View file

@ -6,17 +6,17 @@ from ...typing import AsyncResult, Messages, ImagesType
from ...errors import ResponseError
from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
from .BlackForestLabsFlux1Dev import BlackForestLabsFlux1Dev
from .BlackForestLabsFlux1Schnell import BlackForestLabsFlux1Schnell
from .VoodoohopFlux1Schnell import VoodoohopFlux1Schnell
from .CohereForAI import CohereForAI
from .Janus_Pro_7B import Janus_Pro_7B
from .Phi_4 import Phi_4
from .BlackForestLabs_Flux1Dev import BlackForestLabs_Flux1Dev
from .BlackForestLabs_Flux1Schnell import BlackForestLabs_Flux1Schnell
from .CohereForAI_C4AI_Command import CohereForAI_C4AI_Command
from .DeepseekAI_JanusPro7b import DeepseekAI_JanusPro7b
from .G4F import G4F
from .Microsoft_Phi_4 import Microsoft_Phi_4
from .Qwen_QVQ_72B import Qwen_QVQ_72B
from .Qwen_Qwen_2_5M_Demo import Qwen_Qwen_2_5M_Demo
from .Qwen_Qwen_2_72B_Instruct import Qwen_Qwen_2_72B_Instruct
from .StableDiffusion35Large import StableDiffusion35Large
from .G4F import G4F
from .StabilityAI_SD35Large import StabilityAI_SD35Large
from .Voodoohop_Flux1Schnell import Voodoohop_Flux1Schnell
class HuggingSpace(AsyncGeneratorProvider, ProviderModelMixin):
url = "https://huggingface.co/spaces"
@ -24,20 +24,20 @@ class HuggingSpace(AsyncGeneratorProvider, ProviderModelMixin):
working = True
default_model = Qwen_Qwen_2_72B_Instruct.default_model
default_image_model = BlackForestLabsFlux1Dev.default_model
default_image_model = BlackForestLabs_Flux1Dev.default_model
default_vision_model = Qwen_QVQ_72B.default_model
providers = [
BlackForestLabsFlux1Dev,
BlackForestLabsFlux1Schnell,
VoodoohopFlux1Schnell,
CohereForAI,
Janus_Pro_7B,
Phi_4,
BlackForestLabs_Flux1Dev,
BlackForestLabs_Flux1Schnell,
CohereForAI_C4AI_Command,
DeepseekAI_JanusPro7b,
G4F,
Microsoft_Phi_4,
Qwen_QVQ_72B,
Qwen_Qwen_2_5M_Demo,
Qwen_Qwen_2_72B_Instruct,
StableDiffusion35Large,
G4F
StabilityAI_SD35Large,
Voodoohop_Flux1Schnell,
]
@classmethod
@ -98,3 +98,4 @@ class HuggingSpace(AsyncGeneratorProvider, ProviderModelMixin):
for provider in HuggingSpace.providers:
provider.parent = HuggingSpace.__name__
provider.hf_space = True

1
g4f/gui/client/static/js/chat.v1.js
View file

@ -2145,6 +2145,7 @@ async function on_api() {
+ (provider.vision ? " (Image Upload)" : "")
+ (provider.image ? " (Image Generation)" : "")
+ (provider.nodriver ? " (Browser)" : "")
+ (provider.hf_space ? " (HuggingSpace)" : "")
+ (!provider.nodriver && provider.auth ? " (Auth)" : "");
if (provider.parent)
option.dataset.parent = provider.parent;

1
g4f/gui/server/api.py
View file

@ -67,6 +67,7 @@ class Api:
"image": bool(getattr(provider, "image_models", False)),
"vision": getattr(provider, "default_vision_model", None) is not None,
"nodriver": getattr(provider, "use_nodriver", False),
"hf_space": getattr(provider, "hf_space", False),
"auth": provider.needs_auth,
"login_url": getattr(provider, "login_url", None),
} for provider in __providers__ if provider.working]

119
g4f/models.py
View file

@ -13,14 +13,17 @@ from .Provider import (
Copilot,
DDG,
DeepInfraChat,
Dynaspark,
Free2GPT,
FreeGpt,
HuggingSpace,
G4F,
Janus_Pro_7B,
DeepseekAI_JanusPro7b,
Glider,
Goabror,
ImageLabs,
Jmuz,
LambdaChat,
Liaobots,
OIVSCode,
PerplexityLabs,
@ -28,6 +31,7 @@ from .Provider import (
PollinationsAI,
PollinationsImage,
TeachAnything,
Websim,
Yqcloud,
### Needs Auth ###
@ -88,6 +92,7 @@ default = Model(
Free2GPT,
FreeGpt,
Glider,
Dynaspark,
OpenaiChat,
Jmuz,
Cloudflare,
@ -102,6 +107,7 @@ default_vision = Model(
OIVSCode,
DeepInfraChat,
PollinationsAI,
Dynaspark,
HuggingSpace,
GeminiPro,
HuggingFaceAPI,
@ -120,27 +126,27 @@ default_vision = Model(
gpt_4 = Model(
name = 'gpt-4',
base_provider = 'OpenAI',
best_provider = IterListProvider([DDG, Jmuz, ChatGptEs, PollinationsAI, Yqcloud, Copilot, OpenaiChat, Liaobots])
best_provider = IterListProvider([DDG, Jmuz, ChatGptEs, PollinationsAI, Yqcloud, Goabror, Copilot, OpenaiChat, Liaobots])
)
# gpt-4o
gpt_4o = VisionModel(
name = 'gpt-4o',
base_provider = 'OpenAI',
best_provider = IterListProvider([Blackbox, Jmuz, ChatGptEs, PollinationsAI, Copilot, Liaobots, OpenaiChat])
best_provider = IterListProvider([Blackbox, Jmuz, ChatGptEs, PollinationsAI, Liaobots, OpenaiChat])
)
gpt_4o_mini = Model(
name = 'gpt-4o-mini',
base_provider = 'OpenAI',
best_provider = IterListProvider([DDG, ChatGptEs, Jmuz, PollinationsAI, OIVSCode, Liaobots, OpenaiChat])
best_provider = IterListProvider([DDG, Blackbox, ChatGptEs, Jmuz, PollinationsAI, OIVSCode, Liaobots, OpenaiChat])
)
# o1
o1 = Model(
name = 'o1',
base_provider = 'OpenAI',
best_provider = IterListProvider([Blackbox, OpenaiAccount])
best_provider = IterListProvider([Blackbox, Copilot, OpenaiAccount])
)
o1_mini = Model(
@ -193,19 +199,19 @@ llama_3_70b = Model(
llama_3_1_8b = Model(
name = "llama-3.1-8b",
base_provider = "Meta Llama",
best_provider = IterListProvider([Blackbox, DeepInfraChat, Glider, Jmuz, PollinationsAI, Cloudflare])
best_provider = IterListProvider([DeepInfraChat, Glider, PollinationsAI, AllenAI, Jmuz, Cloudflare])
)
llama_3_1_70b = Model(
name = "llama-3.1-70b",
base_provider = "Meta Llama",
best_provider = IterListProvider([Blackbox, Glider, Jmuz])
best_provider = IterListProvider([Glider, AllenAI, Jmuz])
)
llama_3_1_405b = Model(
name = "llama-3.1-405b",
base_provider = "Meta Llama",
best_provider = IterListProvider([Blackbox, Jmuz])
best_provider = IterListProvider([AllenAI, Jmuz])
)
# llama 3.2
@ -243,7 +249,7 @@ llama_3_2_90b = Model(
llama_3_3_70b = Model(
name = "llama-3.3-70b",
base_provider = "Meta Llama",
best_provider = IterListProvider([DDG, Blackbox, DeepInfraChat, PollinationsAI, Jmuz, HuggingChat, HuggingFace])
best_provider = IterListProvider([DDG, DeepInfraChat, LambdaChat, PollinationsAI, Jmuz, HuggingChat, HuggingFace])
)
### Mistral ###
@ -267,20 +273,14 @@ mistral_nemo = Model(
mixtral_small_24b = Model(
name = "mixtral-small-24b",
base_provider = "Mistral",
best_provider = DDG
)
mixtral_small_28b = Model(
name = "mixtral-small-28b",
base_provider = "Mistral",
best_provider = IterListProvider([Blackbox, DeepInfraChat])
best_provider = IterListProvider([DDG, DeepInfraChat])
)
### NousResearch ###
hermes_2_dpo = Model(
name = "hermes-2-dpo",
hermes_3 = Model(
name = "hermes-3",
base_provider = "NousResearch",
best_provider = Blackbox
best_provider = LambdaChat
)
### Microsoft ###
@ -329,20 +329,20 @@ gemini_exp = Model(
gemini_1_5_flash = Model(
name = 'gemini-1.5-flash',
base_provider = 'Google DeepMind',
best_provider = IterListProvider([Blackbox, Free2GPT, FreeGpt, TeachAnything, Jmuz, GeminiPro])
best_provider = IterListProvider([Free2GPT, FreeGpt, TeachAnything, Websim, Dynaspark, Jmuz, GeminiPro])
)
gemini_1_5_pro = Model(
name = 'gemini-1.5-pro',
base_provider = 'Google DeepMind',
best_provider = IterListProvider([Blackbox, Free2GPT, FreeGpt, TeachAnything, Jmuz, GeminiPro])
best_provider = IterListProvider([Free2GPT, FreeGpt, TeachAnything, Websim, Jmuz, GeminiPro])
)
# gemini-2.0
gemini_2_0_flash = Model(
name = 'gemini-2.0-flash',
base_provider = 'Google DeepMind',
best_provider = IterListProvider([Blackbox, GeminiPro, Liaobots])
best_provider = IterListProvider([Dynaspark, GeminiPro, Liaobots])
)
gemini_2_0_flash_thinking = Model(
@ -412,12 +412,6 @@ blackboxai = Model(
best_provider = Blackbox
)
blackboxai_pro = Model(
name = 'blackboxai-pro',
base_provider = 'Blackbox AI',
best_provider = Blackbox
)
### CohereForAI ###
command_r = Model(
name = 'command-r',
@ -437,6 +431,12 @@ command_r7b = Model(
best_provider = HuggingSpace
)
command_a = Model(
name = 'command-a',
base_provider = 'CohereForAI',
best_provider = HuggingSpace
)
### Qwen ###
qwen_1_5_7b = Model(
name = 'qwen-1.5-7b',
@ -473,7 +473,7 @@ qwen_2_5_1m = Model(
qwq_32b = Model(
name = 'qwq-32b',
base_provider = 'Qwen',
best_provider = IterListProvider([Blackbox, Jmuz, HuggingChat])
best_provider = IterListProvider([Jmuz, HuggingChat])
)
qvq_72b = VisionModel(
name = 'qvq-72b',
@ -498,19 +498,19 @@ deepseek_chat = Model(
deepseek_v3 = Model(
name = 'deepseek-v3',
base_provider = 'DeepSeek',
best_provider = IterListProvider([Blackbox, DeepInfraChat, OIVSCode, Liaobots])
best_provider = IterListProvider([Blackbox, DeepInfraChat, LambdaChat, OIVSCode, Liaobots])
)
deepseek_r1 = Model(
name = 'deepseek-r1',
base_provider = 'DeepSeek',
best_provider = IterListProvider([Blackbox, DeepInfraChat, Glider, PollinationsAI, Jmuz, Liaobots, HuggingChat, HuggingFace])
best_provider = IterListProvider([Blackbox, DeepInfraChat, Glider, LambdaChat, PollinationsAI, Jmuz, Liaobots, HuggingChat, HuggingFace])
)
janus_pro_7b = VisionModel(
name = Janus_Pro_7B.default_model,
name = DeepseekAI_JanusPro7b.default_model,
base_provider = 'DeepSeek',
best_provider = IterListProvider([Janus_Pro_7B, G4F])
best_provider = IterListProvider([DeepseekAI_JanusPro7b, G4F])
)
### x.ai ###
@ -561,14 +561,14 @@ r1_1776 = Model(
nemotron_70b = Model(
name = 'nemotron-70b',
base_provider = 'Nvidia',
best_provider = IterListProvider([HuggingChat, HuggingFace])
best_provider = IterListProvider([LambdaChat, HuggingChat, HuggingFace])
)
### Databricks ###
dbrx_instruct = Model(
name = 'dbrx-instruct',
base_provider = 'Databricks',
best_provider = IterListProvider([Blackbox, DeepInfraChat])
best_provider = DeepInfraChat
)
### THUDM ###
@ -657,6 +657,12 @@ olmoe_0125 = Model(
best_provider = AllenAI
)
lfm_40b = Model(
name = "lfm-40b",
base_provider = "Liquid AI",
best_provider = LambdaChat
)
### Uncensored AI ###
evil = Model(
name = 'evil',
@ -686,7 +692,7 @@ sd_3_5 = ImageModel(
flux = ImageModel(
name = 'flux',
base_provider = 'Black Forest Labs',
best_provider = IterListProvider([Blackbox, PollinationsImage, HuggingSpace])
best_provider = IterListProvider([Blackbox, PollinationsImage, Websim, HuggingSpace])
)
flux_pro = ImageModel(
@ -778,10 +784,9 @@ class ModelUtils:
mixtral_8x22b.name: mixtral_8x22b,
mistral_nemo.name: mistral_nemo,
mixtral_small_24b.name: mixtral_small_24b,
mixtral_small_28b.name: mixtral_small_28b,
### NousResearch ###
hermes_2_dpo.name: hermes_2_dpo,
hermes_3.name: hermes_3,
### Microsoft ###
# phi
@ -821,12 +826,12 @@ class ModelUtils:
### Blackbox AI ###
blackboxai.name: blackboxai,
blackboxai_pro.name: blackboxai_pro,
### CohereForAI ###
command_r.name: command_r,
command_r_plus.name: command_r_plus,
command_r7b.name: command_r7b,
command_a.name: command_a,
### GigaChat ###
gigachat.name: gigachat,
@ -861,19 +866,33 @@ class ModelUtils:
deepseek_v3.name: deepseek_v3,
deepseek_r1.name: deepseek_r1,
nemotron_70b.name: nemotron_70b, ### Nvidia ###
dbrx_instruct.name: dbrx_instruct, ### Databricks ###
glm_4.name: glm_4, ### THUDM ###
mini_max.name: mini_max, ## MiniMax ###
yi_34b.name: yi_34b, ## 01-ai ###
### Nvidia ###
nemotron_70b.name: nemotron_70b,
### Databricks ###
dbrx_instruct.name: dbrx_instruct,
### THUDM ###
glm_4.name: glm_4,
## MiniMax ###
mini_max.name: mini_max,
## 01-ai ###
yi_34b.name: yi_34b,
### Cognitive Computations ###
dolphin_2_6.name: dolphin_2_6,
dolphin_2_9.name: dolphin_2_9,
airoboros_70b.name: airoboros_70b, ### DeepInfra ###
lzlv_70b.name: lzlv_70b, ### Lizpreciatior ###
minicpm_2_5.name: minicpm_2_5, ### OpenBMB ###
### DeepInfra ###
airoboros_70b.name: airoboros_70b,
### Lizpreciatior ###
lzlv_70b.name: lzlv_70b,
### OpenBMB ###
minicpm_2_5.name: minicpm_2_5,
### Ai2 ###
tulu_3_405b.name: tulu_3_405b,
@ -882,7 +901,11 @@ class ModelUtils:
tulu_3_70b.name: tulu_3_70b,
olmoe_0125.name: olmoe_0125,
evil.name: evil, ### Uncensored AI ###
### Liquid AI ###
lfm_40b.name: lfm_40b,
### Uncensored AI ###
evil.name: evil,
#############
### Image ###

30
g4f/tools/run_tools.py
View file

@ -157,6 +157,33 @@ class ThinkingProcessor:
if not start_time and "<think>" not in chunk and "</think>" not in chunk:
return 0, [chunk]
# Handle case where both opening and closing tags are in the same chunk
if "<think>" in chunk and "</think>" in chunk:
parts = chunk.split("<think>", 1)
before_think = parts[0]
thinking_and_after = parts[1]
thinking_parts = thinking_and_after.split("</think>", 1)
thinking_content = thinking_parts[0]
after_think = thinking_parts[1] if len(thinking_parts) > 1 else ""
if before_think.strip():
results.append(before_think)
# Only add thinking content if it contains non-whitespace
if thinking_content.strip():
results.append(Reasoning(thinking_content))
thinking_duration = time.time() - start_time if start_time > 0 else 0
status = f"Thought for {thinking_duration:.2f}s" if thinking_duration > 1 else "Finished"
results.append(Reasoning(status=status, is_thinking="</think>"))
# Important: Add the content that comes after the thinking tags
if after_think.strip():
results.append(after_think)
return 0, results
# Handle thinking start
if "<think>" in chunk and "`<think>`" not in chunk:
before_think, *after = chunk.split("<think>", 1)
@ -183,7 +210,8 @@ class ThinkingProcessor:
status = f"Thought for {thinking_duration:.2f}s" if thinking_duration > 1 else "Finished"
results.append(Reasoning(status=status, is_thinking="</think>"))
if after and after[0]:
# Make sure to handle text after the closing tag
if after and after[0].strip():
results.append(after[0])
return 0, results