Integrations
Works with the models and tools you already use.
If your app can call OpenAI, it can call Behest. Keep your stack, and put every AI call under one control plane.
The short version
Which LLM providers does Behest support?
Behest routes to every major LLM provider through one OpenAI-compatible API: OpenAI, Anthropic, Gemini, and others. Bring your own provider keys or use models through Behest, with cost-aware routing across them.
Get the full integrations guide
Models from OpenAI, Anthropic, Gemini, and others; every major web, mobile, and backend framework; Kubernetes, Docker, Helm, and your cloud; Supabase, Auth0, Clerk, Firebase, and custom sign-in.
What you get with every integration
Cost control on every call
Every model call is tracked to the user, project, and session, and spend limits stop overruns before the provider invoice arrives.
AI Token FinOpsGovernance on every call
On the Enterprise plan, private data is scrubbed and risky prompts are screened before the model sees them, and only the models you approve get used.
AI GovernanceOne place to see it all
Every model, app, and team that runs through Behest shows up in one view, whichever provider or framework is behind it.
The platform
Frequently asked questions
- Do I need to install a library or change my code to use Behest?
- No special library is required. Behest uses the same request format as OpenAI (POST /v1/chat/completions), so if your app can already call OpenAI, it can call Behest by pointing at your Behest URL and API key. It is standard HTTP with JSON from any language. Official Python and Node.js libraries are coming soon. The docs have a copy-and-paste quickstart.
- Can I bring my own provider API keys?
- Yes. On paid plans you can bring your own keys for OpenAI, Anthropic, Gemini, and others, or use models through Behest with pass-through billing. Behest is a SaaS license, not a markup on your tokens.
- Can I call Behest directly from a web or mobile app?
- Yes. Behest handles the browser-to-AI connection (CORS) so React, Vue, Svelte, Angular, and mobile apps can call it directly, with sign-in checks and usage caps applied to every call, and private-data scrubbing on the Enterprise plan. There is no server of your own to build in between.
- Where does Behest run?
- Behest runs as SaaS in Behest's cloud on every plan. On the Enterprise self-hosted option you can run it in your own environment on any Kubernetes cluster, including GKE and AWS EKS, using the container and deployment tools listed in the integrations guide.
- Which sign-in providers work with Behest?
- Supabase and any custom JWT issuer are live today. Auth0, Clerk, and Firebase Auth are coming soon. Behest verifies your users' tokens so every request is tied to a real user for cost tracking and governance.