mirror of
https://github.com/xtekky/gpt4free.git
synced 2025-12-06 02:30:41 -08:00
f4cd4890d3
* feat: enhance provider support and add PuterJS provider - Add new PuterJS provider with extensive model support and authentication handling - Add three new OIVSCode providers (OIVSCodeSer2, OIVSCodeSer5, OIVSCodeSer0501) - Fix Blackbox provider with improved model handling and session generation - Update model aliases across multiple providers for consistency - Mark DDG provider as not working - Move TypeGPT to not_working directory - Fix model name formatting in DeepInfraChat and other providers (qwen3 → qwen-3) - Add get_model method to LambdaChat and other providers for better model alias handling - Add ModelNotFoundError import to providers that need it - Update model definitions in models.py with new providers and aliases - Fix client/stubs.py to allow arbitrary types in ChatCompletionMessage * Fix conflicts g4f/Provider/needs_auth/Grok.py * fix: update Blackbox provider default settings - Changed parameter to use only the passed value without fallback to 1024 - Set to instead of in request payload * feat: add WeWordle provider with gpt-4 support - Created new WeWordle.py provider file implementing AsyncGeneratorProvider - Added WeWordle class with API endpoint at wewordle.org/gptapi/v1/web/turbo - Set provider properties: working=True, needs_auth=False, supports_stream=True - Configured default_model as 'gpt-4' with retry mechanism for API requests - Implemented URL sanitization logic to handle malformed URLs - Added response parsing for different JSON response formats - Added WeWordle to Provider/__init__.py imports - Added WeWordle to default model providers list in models.py - Added WeWordle to gpt_4 best_provider list in models.py * feat: add DocsBot provider with GPT-4o support - Added new DocsBot.py provider file implementing AsyncGeneratorProvider and ProviderModelMixin - Created Conversation class extending JsonConversation to track conversation state - Implemented create_async_generator method with support for: - Streaming and non-streaming responses - System messages - Message history - Image handling via data URIs - Conversation tracking - Added DocsBot to Provider/__init__.py imports - Added DocsBot to default and default_vision model providers in models.py - Added DocsBot as a provider for gpt_4o model in models.py - Set default_model and vision support to 'gpt-4o' - Implemented API endpoint communication with docsbot.ai * feat: add OpenAIFM provider and update audio model references - Added new OpenAIFM provider in g4f/Provider/audio/OpenAIFM.py for text-to-speech functionality - Updated PollinationsAI.py to rename "gpt-4o-audio" to "gpt-4o-mini-audio" - Added OpenAIFM to audio provider imports in g4f/Provider/audio/__init__.py - Modified save_response_media() in g4f/image/copy_images.py to handle source_url separately from media_url - Added new gpt_4o_mini_tts AudioModel in g4f/models.py with OpenAIFM as best provider - Updated ModelUtils dictionary in models.py to include both gpt_4o_mini_audio and gpt_4o_mini_tts * fix: improve PuterJS provider and add Gemini to best providers - Changed client_id generation in PuterJS from time-based to UUID format - Fixed duplicate json import in PuterJS.py - Added uuid module import in PuterJS.py - Changed host header from "api.puter.com" to "puter.com" - Modified error handling to use Exception instead of RateLimitError - Added Gemini to best_provider list for gemini-2.5-flash model - Added Gemini to best_provider list for gemini-2.5-pro model - Fixed missing newline at end of Gemini.py file --------- Co-authored-by: kqlio67 <kqlio67.noreply.github.com> |
||
|---|---|---|
| .. | ||
| examples | ||
| testing | ||
| tool | ||
| unittest | ||