Vercel treats code generation as assumed and builds its value on carrying generated code into production, running it in isolated Firecracker microVMs and billing only for active processor time to match how agents actually run.

Vercel's rebuilt v0 runs in a sandbox tied to real repositories, wraps work in a Git-based review and deploy workflow, and executes untrusted AI code inside Firecracker microVMs — a strong isolation boundary prompted by real incidents like credentials pasted into prompts. Its Fluid compute bills for active processor time while treating wait time as free, matching agentic workloads that spend much time waiting on a model. ✦ AI generated

Article author · ByteByteGo Newsletter · 2026-08-12 · original ↗

Vercel starts from a different premise where code generation is assumed, and the design deals with carrying the generated code into production. ... The rebuilt version of v0, Vercel's generation product, runs on a sandbox. It is an isolated space for executing code that imports a real GitHub repository and automatically pulls in the project's environment variables and configuration. ... Vercel's rationale about this approach is that AI-assisted building is already happening inside companies, and it has produced real failures. Incidents have been reported, such as credentials pasted into prompts, private data reaching the public internet, and deleted databases, often with the audit trail left empty. Therefore, wrapping code generation in real deployment controls is the right response. ... Every sandbox runs inside a Firecracker microVM, a lightweight virtual machine that isolates untrusted code. The reason for this isolation is that the code an AI wrote is code you have yet to review. Therefore, running it needs a boundary strong enough to contain mistakes. ... Vercel's Fluid compute lets several requests share one running instance, with one using the processor while another waits on input or output. It charges for active processor time while treating wait time as free. Agentic workloads spend much of their time waiting on a model to respond, so this pricing matches the real work being done.

Read full article ↗excerpt · fair-use quotation