hlohaus
4dce265340
fix: update default_model to nemotron-3-super in Ollama class
2026-05-15 14:24:04 +02:00
hlohaus
8defec4e33
fix: ensure filename has .html extension in _playground method
2026-05-15 04:57:18 +02:00
hlohaus
8ba8697e97
feat: add playground route with caching and GitHub integration for dynamic content serving
2026-05-12 20:14:19 +02:00
hlohaus
9b608fcc49
fix: ensure api_key validation checks for string type in is_provider_api_key function
2026-05-03 20:17:05 +02:00
hlohaus
074280ca98
feat: update PuterJS to use gpt-5.1 as default model and streamline model fetching logic
...
feat: enhance Usage class to include prompt and completion token counts
2026-05-03 17:12:06 +02:00
hlohaus
6365699b1a
fix: adjust timeout assertion in security hardening tests to allow for longer execution time
2026-05-03 04:38:38 +02:00
hlohaus
ce42a89240
fix: update get_media_extension to return empty string for unsupported extensions
2026-05-03 02:24:23 +02:00
hlohaus
5c754a4da3
feat: enhance API key handling to include validation for provider-specific keys
2026-05-03 00:45:50 +02:00
hlohaus
da29e7cb7a
feat: update g4f-perplexity provider configuration to use openai-completions API
2026-05-02 09:16:15 +02:00
H Lohaus
7773c8288c
Update PollinationsAudio.py
2026-05-01 14:01:34 +02:00
H Lohaus
a51753a67c
Update __init__.py
2026-05-01 14:00:18 +02:00
hlohaus
b4e6f29164
feat(PollinationsAudio): add new audio provider with transcription and speech capabilities
2026-05-01 12:43:05 +02:00
H Lohaus
2e7a021097
Update __main__.py
2026-04-21 22:42:29 +02:00
H Lohaus
0393f1053f
Merge pull request #3428 from vyguera/fix/openai-chat-gizmo-conversation-mode
...
fix(OpenaiChat): use gizmo_interaction conversation_mode for custom GPTs
2026-04-21 22:41:41 +02:00
H Lohaus
06c1069cc1
Merge pull request #3429 from 3mora2/fix_qwen
...
Fix get_models ,Update Qwen provider models and thinking mode
2026-04-21 22:40:43 +02:00
Ammar
ccb5c070f4
Update Qwen provider models and thinking mode
...
Refresh static model lists and improve dynamic model discovery/parsing. get_models now accepts an optional token, sends headers, and uses safe dict access (model.get(...)) to build text/image/vision/models lists and set default_model. Adjust generation logic so thinking_mode defaults to "Auto" when reasoning is enabled (otherwise "Fast"), introduce an auto_thinking variable, and propagate token headers into the StreamSession.
2026-04-21 10:32:44 +02:00
ruben.viguera
2adec1bd1d
fix(OpenaiChat): use gizmo_interaction conversation_mode for custom GPTs
...
When gizmo_id is passed, conversation_mode was hardcoded to primary_assistant
in both request payload data dicts, causing the custom GPT's system instructions
to be ignored and the base model to respond instead.
Fixes #3427
2026-04-20 09:54:54 +02:00
H Lohaus
01177797e3
Update README.md
2026-04-17 22:20:15 +02:00
hlohaus
cc0ad7a9f8
Refactor provider method handling for improved clarity and consistency; update mocks and add SKILL.md for usage guidance
2026-04-17 14:07:49 +02:00
hlohaus
06df47f279
Refactor provider methods to unify async and sync handling, enhance clarity, and improve error management
2026-04-16 02:38:03 +02:00
H Lohaus
65ee4cb088
Merge pull request #3420 from 3mora2/enhance-Qwen
...
Add thinking_mode and feature_config handling
2026-04-15 19:08:13 +02:00
hlohaus
3b9fb6f5fe
Fix parameter assignment in create_func for clarity
2026-04-07 17:12:20 +02:00
hlohaus
38bdcca86a
Fix provider name assignment in Backend_Api for clarity
2026-04-07 17:05:56 +02:00
hlohaus
dfbdf56ec4
Refactor provider initialization in Backend_Api to enhance clarity
2026-04-07 17:02:09 +02:00
hlohaus
86eab3b91c
Refactor media rendering to yield messages only if content parts exist
2026-04-07 16:28:14 +02:00
hlohaus
97f99539b1
Update Claude provider to use new base URL and enhance cookie handling
2026-04-07 16:08:23 +02:00
hlohaus
6aeb41f803
Remove redundant try block in main function for client mode execution
2026-04-07 12:42:19 +02:00
H Lohaus
a6a26faf47
Merge pull request #3419 from xtekky/copilot/update-finish-reason-tool-call
...
Fix finish_reason to return "tool_calls" when tool calls are present
2026-04-06 14:22:26 +02:00
Ammar
327b698d0b
Add thinking_mode and feature_config handling
...
Introduce a thinking_mode option and build a conditional feature_config payload for Qwen requests: when reasoning is enabled the config includes auto_thinking, thinking_mode, thinking_enabled, output_schema, research_mode and auto_search; otherwise it provides a minimal config with thinking_budget. Also include minor cleanup and formatting/PEP8 fixes (spacing around defaults, blank-line adjustments, and reflowed long request lines) to improve readability.
2026-04-06 13:10:51 +02:00
copilot-swe-agent[bot]
90fb39a47a
Fix finish_reason to return tool_calls when tool calls are present
...
Agent-Logs-Url: https://github.com/xtekky/gpt4free/sessions/c0d4d1fc-0e31-4a63-8aa1-79132a731019
Co-authored-by: hlohaus <983577+hlohaus@users.noreply.github.com>
2026-04-06 08:43:11 +00:00
copilot-swe-agent[bot]
e03f0b7e8f
Initial plan
2026-04-06 08:40:34 +00:00
hlohaus
9f9d18cc4d
Add debug mode to CLI and improve argument parsing error handling
2026-04-06 06:10:08 +02:00
hlohaus
889e09770b
Enhance security in pa_provider by implementing import restrictions for sensitive submodules; remove unused AuthManager import in base_provider
2026-04-06 05:02:23 +02:00
hlohaus
cd8c096f68
Fix chunk handling in Backend_Api to ensure content is properly yielded as a string
2026-04-06 03:06:46 +02:00
H Lohaus
1325c4d9f9
Merge pull request #3418 from xtekky/copilot/add-pa-py-file-parser-executor
...
Refactor PA backend conversation endpoint for improved streaming and …
2026-04-06 00:39:51 +02:00
hlohaus
05a9ac45f4
Refactor PA backend conversation endpoint for improved streaming and error handling; adjust session key generation to use 1024 bits for user input compatibility
2026-04-06 00:36:32 +02:00
H Lohaus
640c45ba0b
Merge pull request #3415 from xtekky/copilot/add-pa-py-file-parser-executor
...
Add .pa.py safe provider executor, workspace file MCP tools, --safe server mode, sandbox security hardening, /pa/* API routes, and secure workspace file serving with origin isolation
2026-04-05 23:37:44 +02:00
hlohaus
b09317fe47
Add PA file handling and secure serving in MCPServer; update FileWriteTool to include file URL
2026-04-05 23:36:52 +02:00
hlohaus
758df77264
Refactor MCPServer response structure and update FileListTool workspace handling for safe mode
2026-04-05 21:30:47 +02:00
hlohaus
e4fcd38d3c
Add default active state to OperaAria and enhance quota tracking in Ollama
2026-04-05 19:52:04 +02:00
H Lohaus
80477446ff
Merge pull request #3417 from xtekky/copilot/fix-thumbnail-creation-error
...
Fix RGBA image thumbnail creation failing with JPEG format
2026-04-05 19:47:26 +02:00
hlohaus
82114693b4
Remove unused login URL from Azure class
2026-04-05 18:42:47 +02:00
hlohaus
d496433e8d
Add quota URL to GeminiPro class for API usage tracking
2026-04-05 18:30:55 +02:00
hlohaus
270da33792
Add Azure quota URL, enhance API model retrieval, and improve error handling for invalid API keys
2026-04-05 13:56:21 +02:00
H Lohaus
349bba9c90
Update README.md
2026-04-05 12:45:16 +02:00
H Lohaus
8211978a7f
Update README.md
2026-04-05 12:38:34 +02:00
copilot-swe-agent[bot]
c406ed38a1
fix: convert RGBA images to RGB before saving as JPEG thumbnail
...
Agent-Logs-Url: https://github.com/xtekky/gpt4free/sessions/0f9b1c1f-a0a4-49a5-9793-b30940de945f
Co-authored-by: hlohaus <983577+hlohaus@users.noreply.github.com>
2026-04-04 23:36:30 +00:00
copilot-swe-agent[bot]
2f00d7ce12
Initial plan
2026-04-04 23:35:35 +00:00
H Lohaus
35b8e96378
Update config.py
2026-04-05 01:07:34 +02:00
H Lohaus
644225096b
Update config.py
2026-04-05 01:04:07 +02:00
