You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Users who run LiteLLM as a proxy for 100+ LLM providers can't point opencode at it. They have to configure each provider individually even when LiteLLM already provides a unified OpenAI-compatible endpoint.
Suggested Fix
Add LiteLLM as a built-in provider using the existing @ai-sdk/openai-compatible adapter, same as other OpenAI-compatible providers in the codebase.
I see this was closed as not planned by github-actions? Can this be revisited, as it's currently a massive undertaking to manually sync all models and their related metadata/capabilities.
Activity
github-actions commentedon Apr 12, 2026
This issue doesn't fully meet our contributing guidelines.
What needs to be fixed:
[FEATURE]:(current title:feat: add LiteLLM provider support)Please edit this issue to address the above within 2 hours, or it will be automatically closed.
This issue might be a duplicate of existing issues. Please check:
autoloadoption #13891: Feature request to add LiteLLM proxy support with auto-loading models vialitellmProxyandautoloadoptions — covers very similar ground (built-in LiteLLM provider integration)If you believe this was flagged incorrectly, please let a maintainer know.
github-actions commentedon Apr 13, 2026
This issue has been automatically closed because it was not updated to meet our contributing guidelines within the 2-hour window.
Feel free to open a new issue that follows our issue templates.
Lukas-Ertl commentedon Jun 24, 2026
+1
jsilvermist commentedon Sep 13, 2026
I see this was closed as
not plannedby github-actions? Can this be revisited, as it's currently a massive undertaking to manually sync all models and their related metadata/capabilities.