Changelog

Fast mode on AI Gateway routes a request to a model's faster serving path for lower latency, set with the gateway.speed option or a fast slug, with automatic fallback to standard speed and per-request speed billing.

Backend metrics like database query time will reach the browser's Performance API on August 10, 2026, when the CDN stops stripping the header.

The @vercel/sandbox SDK now supports multiple Linux users and groups, so you can run agents side by side in a single Sandbox.

MiniMax H3 (Hailuo 03) is now available on AI Gateway, a video generation model that produces 2K video from a text prompt, an image, or reference material.

Inkling-Small from Thinking Machines is now available on AI Gateway, an efficient open-weights multimodal Mixture-of-Experts model with a 1M token context window and controllable thinking effort.

Configure OIDC policies for your team's CI/CD workflows to access Vercel Remote Cache using Turborepo.

GPT-5.6 Luna is 80% cheaper, GPT-5.6 Terra is 20% cheaper, and GPT-5.6 Sol's fast mode is now 2.5x faster on AI Gateway, with no markup on token pricing.

We made deployments up to 7 seconds faster end to end. Every build sheds up to 5 seconds of fixed platform overhead, and the latest Vercel CLI saves up to 2 more.