APIs, MCP servers and models to host open models
9 tools in the catalog do this, by name.
Amazon BedrockProviderAWS service for calling 100+ foundation models through one API
falModelHosted API for 1,000+ open-source and commercial image, video and audio models
Fireworks AIProviderInference, fine-tuning and dedicated GPU deployments for open models
GroqProviderLow-latency inference API for open LLMs and Whisper
LocalAIPackageRun LLM, image and audio models on your own hardware behind OpenAI-style APIs
ModalPlatformServerless cloud for running Python on GPUs, billed per second
ReplicateModelCloud API for running and fine-tuning AI models
Together AIProviderServerless and dedicated inference, fine-tuning and GPUs for open models
vLLMPackageOpen-source engine for serving LLMs with high throughput on GPUs