feat: Add DeepSeek as supported LLM provider #32
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode characters
- Add DEEPSEEK to LLMVendor enum - Update API key pattern detection for DeepSeek keys - Add deepseek-chat as default model for DeepSeek - Configure LLM client to properly handle DeepSeek API - Update README with DeepSeek as supported provider - Add tests for DeepSeek integration
OtherVibes requested changes Sep 25, 2025
There was a problem hiding this comment.
Hi @groxaxo 👋
Please focus on LLM_BASE_URL - it will be a great enhancement for this project
Thanks,
Zvi and the OtherVibes team
There was a problem hiding this comment.
@groxaxo - please exclude this file from the PR
| LLMVendor.GROQ: re.compile(r"^gsk_[a-zA-Z0-9]{50,}"), | ||
| LLMVendor.XAI: re.compile(r"^xai-[a-zA-Z0-9]{40,}"), | ||
| LLMVendor.OPENROUTER: re.compile(r"^sk-or-[a-zA-Z0-9_-]{48}"), | ||
| LLMVendor.DEEPSEEK: re.compile(r"^sk-[a-zA-Z0-9]{20,}-deeplearning-ai$"), # DeepSeek API keys end with -deeplearning-ai |
There was a problem hiding this comment.
this regex does not seem to be valid for DeepSeek keys.
| self._litellm.aws_access_key_id = api_key | ||
| elif vendor == LLMVendor.VERTEX_AI: | ||
| # Vertex AI uses service account JSON | ||
| self._litellm.vertex_ai_key = api_key |
There was a problem hiding this comment.
As written in the note, DeepSeek are OpenAI compatible.
It would make much more sense to refactor this PR to add support to LLM_BASE_URL so users can use DeepSeek using customized URL - this would be excellent contribution.
| @@ -0,0 +1,44 @@ | |||
| """ | |||
There was a problem hiding this comment.
Tests should be written with pytest
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit. This suggestion is invalid because no changes were made to the code. Suggestions cannot be applied while the pull request is closed. Suggestions cannot be applied while viewing a subset of changes. Only one suggestion per line can be applied in a batch. Add this suggestion to a batch that can be applied as a single commit. Applying suggestions on deleted lines is not supported. You must change the existing code in this line in order to create a valid suggestion. Outdated suggestions cannot be applied. This suggestion has been applied or marked resolved. Suggestions cannot be applied from pending reviews. Suggestions cannot be applied on multi-line comments. Suggestions cannot be applied while the pull request is queued to merge. Suggestion cannot be applied right now. Please check back later.
This PR adds DeepSeek as a supported LLM provider for MCP as a Judge.\n\n## Summary of Changes\n- Add DEEPSEEK to LLMVendor enum\n- Add API key pattern detection for DeepSeek (keys ending with -deeplearning-ai)\n- Add deepseek-chat as default model for DeepSeek\n- Configure LLM client to properly handle DeepSeek API calls\n- Update README to include DeepSeek as a supported provider\n- Add comprehensive tests for DeepSeek integration\n\n## Testing\n- All existing tests continue to pass (179/179)\n- New DeepSeek-specific tests added (4/4 passing)\n- Integration verified with mock API key patterns\n\nThis allows users to configure MCP as a Judge with DeepSeek API keys, providing another powerful LLM option for code evaluation and review.