FreeLLMAPI is an OpenAI-compatible proxy for routing requests across free tiers of multiple LLM providers through one /v1 endpoint, with failover, encrypted keys, and an admin dashboard.
FreeLLMAPI is an OpenAI-compatible proxy for routing requests across the free tiers of multiple LLM providers behind a single /v1 endpoint. The project positions itself as a way to combine individual free plans into one shared inference surface for personal experimentation.
It supports a long list of provider integrations, plus any custom OpenAI-compatible endpoint such as llama.cpp, LM Studio, vLLM, or a local Ollama instance. The proxy handles model routing, automatic failover, encrypted storage for upstream keys, and a dashboard for managing keys and reviewing usage.
Aggregates the free tiers of providers including Google, Groq, Cerebras, NVIDIA, Mistral, OpenRouter, GitHub Models, Cohere, Cloudflare, HuggingFace, Z.ai, Ollama, Kilo, Pollinations, LLM7, OVH AI Endpoints, and OpenCode Zen behind one OpenAI-compatible surface.
Uses a router that selects a model for each request, falls back to the next provider when one is rate-limited, returns 429/5xx, or times out, and keeps short cooldowns for failed keys.
Tracks RPM, RPD, TPM, and TPD per provider, model, and key, and keeps sticky sessions on the same model for about 30 minutes during multi-turn conversations.
Stores provider API keys encrypted with AES-256-GCM in SQLite, while clients authenticate to the proxy with a single unified bearer token.
Exposes /v1/chat/completions, /v1/models, /v1/responses, /v1/embeddings, streaming, non-streaming, and OpenAI-style tool calling for compatible clients.
Includes a React + Vite admin dashboard for managing keys, ordering fallback chains, viewing analytics, and running prompts in a playground.
Point an OpenAI SDK, LangChain, LlamaIndex, Continue, or similar client at the proxy and keep the same application code while swapping the upstream path to /v1.
Add provider keys for several free-tier services and let the router choose an available model, then fail over automatically when one provider is throttled or unavailable.
Run the Docker Compose setup locally or on a small server to keep the API, dashboard, and SQLite data in one self-hosted environment.
Use the admin dashboard to reorder fallback chains, inspect latency and token usage, and test prompts before wiring a client into the proxy.
Connect a custom OpenAI-compatible backend such as LM Studio, llama.cpp, vLLM, or local Ollama through the same unified router.
FreeLLMAPI is designed to work with OpenAI-compatible clients. The README says you can point any OpenAI SDK or compatible client such as LangChain, LlamaIndex, Continue, or Hermes at the proxy by changing the base URL.
The Docker guide says Docker Compose is the recommended way to run it for personal use. It serves the API and dashboard from one process on port 3001, with SQLite persisted in a named volume.
The README says the proxy implements /v1/chat/completions, /v1/models, /v1/responses, /v1/embeddings, streaming and non-streaming responses, and OpenAI-style tool calling. It does not implement image generation, audio, legacy completions, moderation, multiple completions per request, or per-user billing.
The project is built around a single-user setup. The README explicitly says per-user billing and multi-tenant auth are not supported yet.
AakarDev AI helps teams manage AI provider access, project-level setups, logs, and analytics from one dashboard. It supports BYOK workflows and lists providers including OpenAI, Google Gemini, Anthropic, Groq, Mistral AI, and Perplexity AI.
BookAI allows you to chat with your books using AI by simply providing the title and author.
Skills Janitor is a GitHub-hosted set of slash commands for auditing, tracking, and managing Claude Code and OpenAI Codex skills. It helps users find duplicates, broken links, and unused skills, then clean them up with self-contained commands.
FeelFish is a PC client for AI-assisted novel writing, designed to help fiction writers plan characters and settings, draft and revise long-form content, and manage story context. It includes a free tier and paid plans, with support for multiple large-model providers.
Benchspan is an AI agent security platform that discovers agents, blocks prompt injection and data exfiltration in real time, and supports pre-launch red teaming. It is aimed at teams running agents in production and includes Python and TypeScript SDKs.
ChatBA is a generative AI product for creating slide decks from prompts. The public site emphasizes instant presentation generation and includes help content for templates, sharing, and data sources.
