Discofork
Repository brief

BerriAI/litellm

Read the upstream summary on the left, browse the cached forks below it, and load each fork comparison into the right-hand panel.

Cached analysis
cached 2026-03-30T19:56:26.131Z
1mo ago

BerriAI/litellm

LiteLLM is a large, actively maintained Python SDK and AI gateway for calling 100+ LLM APIs through an OpenAI-style interface. It also includes a proxy server, cost tracking, guardrails, load balancing, logging, and enterprise-oriented deployment/docs around the proxy.

GitHub
Loading tags...
Stars41,563
Forks6,857
Default branchmain
Last pushed2026-03-30T18:42:15Z
Recommended shortcuts

Jump straight into Discofork's strongest cached fork picks, or open a compare view in one click.

Forks

Choose a fork to inspect

9 of 9 fork briefs
Selected

Prefer upstream unless you specifically need this exact snapshot. This fork does not add capabilities, and it is far enough behind that it likely misses important fixes and features.

Prefer upstream unless you specifically need this fork’s repository identity or want a close mirror to build on. It offers no visible added capability and is already 11 commits behind, so adopters should expect upstream to be safer and more complete.

Treat this as a stale snapshot of LiteLLM, not a maintained fork. Prefer upstream unless you specifically need this exact historical state and can own the maintenance burden yourself.

Prefer upstream unless you specifically need this fork's older, customized proxy/UI behavior and are willing to maintain a large backlog of merges yourself. For new adopters, the fork looks too stale to be a safe default.

Choose this fork only if its specific UI/admin and guardrail tweaks match your deployment and you are prepared to maintain a large upstream gap. If you want current provider support, proxy fixes, and lower operational risk, upstream is the safer choice.

Prefer upstream LiteLLM unless you specifically need an older pinned snapshot; this fork shows no added capabilities and is far behind current upstream.

Prefer upstream unless you specifically need the fork’s local proxy/auth or deployment tweaks and are prepared to own a large maintenance burden. For most adopters, the fork is too stale and too far behind to be a safe default.

Choose this fork only if YandexGPT support is the main requirement and you are prepared to maintain a heavily diverged, stale codebase. If you want the current LiteLLM feature set, upstream is the safer choice.

Choose this fork only if its Responses/streaming fixes match a current blocker. If you need broad LiteLLM coverage, active parity, or the full proxy/platform surface, upstream is the safer choice.