ToolsThe story, in brief
AI Gateway is now generally available
Not a pilot. Vercel's AI Gateway now routes to 100+ models with sub-20ms latency and zero token markup.
The KeyNews take
Why it matters
Vercel democratizes multi-model inference by removing pricing friction and operational complexity. For founders building AI products, this eliminates vendor lock-in while simplifying cost tracking and failover—critical for production deployments.
The key facts
7 to knowAI Gateway now generally available
Sub-20ms latency routing across multiple inference providers
Transparent pricing with no markup on tokens
Supports Bring Your Own Keys (BYOK)
Automatic failover for higher availability
Works with Vercel AI SDK and OpenAI-compatible endpoints
Single model string switch for implementation
Go to the source
Vercel Blogvercel.com
Publisher excerpt: is now generally available, providing a single unified API to access hundreds of AI models with transparent pricing and built-in observability.AI Gateway With sub-20ms latency routing across multiple inference providers, AI Gateway delivers: You can use AI Gateway with the or through the…