Vercel Ships Workflows GA, Feature Flags, Elastic Builds, and AI Gateway Privacy Controls

A New Programming Model for Durable Execution: Vercel Workflows Goes GA

TLDR: Vercel Workflows is now generally available after processing over 100 million runs in beta. It's a programming model that puts orchestration directly in your application code, eliminating the need for separate queue workers and orchestration services.

A new programming model for durable execution

Vercel Flags Is Now Generally Available

TLDR: Vercel Flags is a feature flag system built directly into the Vercel platform, now generally available with Next.js and SvelteKit framework integrations and OpenFeature standard support.

Vercel Flags is now generally available

Elastic Build Machines Is Now GA

TLDR: Vercel's Elastic Build Machines assign the right compute size to each project individually rather than using a one-size-fits-all build configuration. 80% of projects see cost reductions, 20% get auto-upgraded for faster builds.

Elastic Build Machines is now GA

Team-Wide Zero Data Retention and Prompt Training Controls on AI Gateway

TLDR: Vercel's AI Gateway now supports Zero Data Retention at the team level, enforcing that no request data is stored by any AI provider across your entire team's traffic without requiring per-request configuration.

Team-wide Zero Data Retention and prompt training controls now on AI Gateway

Base Fee for Observability Plus Removed

TLDR: Vercel has dropped the $10 base fee for Observability Plus. You now pay only for the observability events you actually collect, with no floor charge.

Base fee for Observability Plus removed

Logs Filtering for Vercel Workflows Now Available

TLDR: You can now filter Vercel logs by Workflow Run ID and Workflow Step ID, making it possible to see all logs for a specific workflow run in one view instead of hunting across individual request logs.

Logs filtering for Vercel Workflows now available

Vercel Sandbox Now Supports Up to 32 vCPU and 64 GB RAM

TLDR: Vercel Sandbox has expanded its resource ceiling to 32 vCPUs and 64 GB RAM for Enterprise customers, enabling resource-intensive workloads that weren't previously possible in sandboxed environments.

Vercel Sandbox now supports up to 32 vCPU + 64 GB RAM configurations