Vercel AI Gateway
One API key and endpoint for hundreds of AI models.
About
Vercel AI Gateway is a unified API that routes AI SDK calls to hundreds of models through a single endpoint and credential. Instead of wiring a separate SDK and API key per provider, developers name the model inline as a "provider/model" string and the gateway handles routing, authentication, automatic failover, cost tracking, and observability. It reached general availability in August 2025 and is the default inference path for the Vercel AI SDK.
Pricing is provider list price on prepaid credits with zero markup on tokens, including bring-your-own-key on paid tiers, plus zero data retention by default. Honest caveats: it is managed-only with no self-host option, governance features are thin (no gateway-side semantic cache or built-in guardrails), and add-ons like team-wide zero-data-retention are billed per request.
What it can do
- Single OpenAI-compatible endpoint for hundreds of models
- Automatic provider failover and per-request cost/latency tracking
- Zero data retention and BYOK support on paid tiers
Setup
Create an AI Gateway API key in the Vercel dashboard and call https://ai-gateway.vercel.sh/v1.
Prerequisites
A Vercel account; an app using the AI SDK or any OpenAI-compatible client.
Use cases
Works with
Trust score — 9/10
Scored 9/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 78; Trending score 86/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.