LLM APIs
Hosted language models and the gateways that route between them.
JSON: /api/v1/directory.json (filter on category=llm-apis)
OpenAI API
Hosted GPT-family language, image, speech and embedding models behind a single REST API, with function calling and structured outputs.
PaidAPI keyAnthropic API
Access to the Claude model family over REST, with tool use, long context windows and prompt caching.
PaidAPI keyGoogle Gemini API
Google's multimodal Gemini models via API key, with a free tier for prototyping and support for large context inputs.
Free tierAPI keyGroq
Very fast inference for open-weight models on custom hardware, exposed through an OpenAI-compatible endpoint.
Free tierAPI keyOpenRouter
One API and one bill for hundreds of models from many providers, with routing and fallback between them.
PaidAPI keyMistral AI
European lab offering open and commercial chat, code and embedding models through a hosted API.
Free tierAPI keyCohere
Enterprise-oriented language models with strong embedding and reranking endpoints for retrieval pipelines.
Free tierAPI keyTogether AI
Serverless and dedicated inference for open-source models, plus fine-tuning and image generation.
PaidAPI keyFireworks AI
Fast hosted inference for open models with function calling, JSON mode and fine-tuning support.
PaidAPI keyHugging Face Inference Providers
Unified access to models from the Hugging Face Hub, served by several inference partners behind one token.
Free tierAPI keyPerplexity Sonar API
Search-grounded answer models that return responses with citations, useful when an agent needs current facts.
PaidAPI keyDeepSeek API
Low-cost chat and reasoning models served through an OpenAI-compatible interface.
PaidAPI keyxAI API
Grok models for chat, reasoning and vision, available through an OpenAI-compatible API.
PaidAPI keyReplicate
Run community and official models, mostly image, video and audio, with a simple prediction API and per-second billing.
PaidAPI keyCloudflare Workers AI
Run open models on Cloudflare's network from a Worker or REST call, with a daily free allowance.
Free tierAPI keyOllama
Run open-weight models locally and talk to them over a simple HTTP API on your own machine.
FreeNo auth