#litellm

1 place to call all your agents - OpenCode, Hermes, Claude Managed Agents, Cursor Agents API, DeepAgents.

Prism Reviewer is an agentic, AI-driven multi-agent code review system orchestrated via LangGraph and LiteLLM. It acts as an autonomous gatekeeper for pull requests by performing targeted static analysis, dependency scanning, AST-based symbol inspection, and parallel LLM-guided code evaluation.

Gateway-agnostic control plane for LLMs — an OpenAI-compatible governance proxy (auth, guardrails, budgets, routing, cost, observability) over a swappable engine: Bifrost / LiteLLM / Portkey / Direct. Own the control plane, swap the translator.

Windows 一键脚本：把 Claude Code 路由到任意 AI 提供商。默认 DeepSeek 直连，支持 Anthropic/OpenAI 两种协议（MiMo/Gemini/通义/Kimi 等），自动装依赖、小白友好。

Deterministic health monitoring for LLM conversations. Detects context pressure, drift, repetition and tool anomalies without adding anything to model context.

Run-scoped LLM observability control plane backed by LiteLLM

Flagship free FastAPI starter for LLM routing/cost visibility; $29 Gumroad kit adds deployment, hardening, routing, and verification guides.

A lightweight, embeddable web chat widget powered by LLMs. Easily integrate an AI chatbot into any website with a single script tag.

🌱 Self-growing task tree for LLM agents — every agent decides its own split. Recursive, not predefined.

Monk LLM Gateway: one API in front of every model. An LLM gateway and router built from scratch (proxy, routing, retry/fallback, caching, budgets), with an interactive course.

Daily 0-100 macro tailwind/headwind index per asset from the economic calendar, RSS news and LLM-extracted institutional forecasts. Python, SQLite, LiteLLM.

Local-model gateway for Claude Code, Zoo Code, and Kilo: route AI coding agents through llama.cpp/Ollama/vLLM on Distrobox. No API keys, no cloud fallback, nothing auto-starts - local mode either works or fails cleanly. Classic LiteLLM proxy, or the optional single-provider Kilo mode.

Self-hosted routing proxy for Claude Code and any Anthropic-format client: nine modes across Anthropic, MiniMax, GLM/z.ai, Qwen and local models. One endpoint, per-chat isolation, no IDE restarts.

Unified LLM client — one interface for OpenAI, Anthropic, Gemini, Groq, DeepSeek. Parallel calls, vision support, cost tracking. Powers RawEval's multi-model evaluation pipeline.

🤷🏾‍♀️ LiteLLM Router Chat is a FastAPI and LiteLLM-based chatbot automatically retries failures and switches between Groq and Gemini models based on availability. I am also hoping to separate model registry, tool registry, and app code, making it model provider-agnostic (in master).