Sorting LiteLLM GitHub issues by difficulty to implement. Higher = harder.

{
The left issue is harder because it requires tracing a version-specific regression across budget accounting, enforcement paths, deployment conditions, and authorization behavior, with regression tests to prevent further bypasses. The right issue is a comparatively bounded provider-integration task using existing OpenAI-compatible routing, documentation, and focused tests.
}
https://github.com/berriai/litellm/issues/26672 3:1 https://github.com/berriai/litellm/issues/34858