Vercel Functions on Node.js and Python runtimes now support execution durations up to 30 minutes for Pro and Enterprise teams, up from the previous 800-second ceiling. The feature is in beta, requires Fluid Compute to be enabled, and is configured via the maxDuration setting in your route file or vercel.json.
The billing model is the reason this matters beyond the raw time limit. Fluid Compute only charges for active CPU time, pausing the clock during I/O waits like LLM calls, database queries, and third-party API responses. Long-running workloads like multi-minute streaming AI responses, OCR pipelines, browser automation, and document processing become viable without ballooning costs.
Support for additional runtimes is listed as coming soon, and the full documentation covers exactly how duration interacts with Fluid Compute pricing. If you are running Workflow steps or Queue handlers that have been hitting timeouts, that configuration detail is worth reading closely.
[READ ORIGINAL →]