Together AI
Inference API with 200+ open-source models at competitive speeds. Popular for running Llama, Mistral, and other open models at scale.
OpenAI API
API access to GPT-5, GPT-5.5, o3/o4 reasoning models, and the Responses API; plus embeddings, image, audio, and Realtime endpoints. The most widely deployed LLM API in production.