One endpoint, every model.
Rota Engine normalizes every major provider behind a single OpenAI-compatible endpoint. Routing, failover and budgets are policy decisions — not code changes.
OpenAI
GPT, o-series reasoning and embedding models — the full catalog, always current.
Anthropic
The complete Claude family, from fast tiers to frontier reasoning.
Google Gemini
Every Gemini model, multimodal and reasoning tiers included.
xAI
The full Grok family, live the day xAI ships it.
Azure OpenAI
The OpenAI catalog, deployed in your region with private networking.
AWS Bedrock
Claude, Llama, Titan — every model Bedrock hosts, one key.
Vertex AI
Gemini and every model Google Cloud serves through Vertex.
Mistral
Every Mistral model, open-weight and commercial.
Cohere
Command and Embed — generation and retrieval, unified.
Groq
Ultra-fast inference for every open-weight model Groq serves.
Together / Fireworks
Hundreds of open-weight models, one governed endpoint.
A new model ships.You already support it.
Rota Engine isn't a fixed list — it's a protocol. Any model any provider exposes through an OpenAI-compatible or native API is already routable, policy-governed and observable. If it can serve a token, Rota already speaks to it.
Need a provider that isn't listed?
Custom adapters ship in days. Bring your endpoint spec and we wire it into the same policy plane.