Skip to main contentFAQ
Is the API OpenAI‑compatible?
Yes. LLMGrid uses OpenAI‑style endpoints/objects so your code ports cleanly.
Can I swap models without client changes?
Yes; update your route and policies, keep the same endpoint.
Do you support streaming?
Yes — SSE for /chat/completions is supported. (Standard gateway behavior.)
How do I control costs?
Budgets, rate limits, per‑user caps, alerts, and policy enforcement (multi‑tenant hierarchy).
How are tokens metered?
Per tenant/project/user/team with dashboards and exportable usage. 
