Vercel makes Python generally available and launches Daemons in beta

Python moves to general availability

According to Vercel, developers can now build backends in Python with the same infrastructure and workflow they use for frontends, including local development, previews and production deployments. Frameworks such as FastAPI, Flask and Django run on standard CPython with PyPI dependencies, while Vercel handles framework detection, builds and scaling.

Vercel says 150,000 teams already run Python backends on its platform, including engineers at OpenAI, Polymarket and Exa. Python SDKs let applications use Vercel products such as Vercel Blob, Vercel Queues, Vercel Workflows, Vercel Sandbox and the AI SDK for Python.

Delphi, which builds AI digital minds from books and talks, moved its Python backend from AWS ECS to Vercel. According to Vercel, the team has ten engineers and no dedicated infrastructure role, ships to production more than 100 times a day and uses Queues for background jobs. Founding Engineer Spencer Schoeben is quoted as saying: "With Vercel, we barely think about infrastructure at all."

Containers for applications in any language

The Vercel container runtime lets teams deploy applications written in any language, packaged with the OS libraries and executables they need. Vercel says that since container support entered Public Beta, 5,000 teams have shipped 150,000 container deployments.

Containers deploy as Vercel Functions on Fluid compute and scale with traffic. They use Active CPU pricing, which according to Vercel does not charge for idle time spent waiting for API responses, database queries or LLM inferencing. When deploying through Git or the Vercel CLI, Vercel builds the image and stores it in the Vercel Container Registry.

Vercel Services allows a backend to be developed and deployed alongside a frontend. Vercel CDN serves traffic on custom domains, Vercel WAF protects it, logs help investigate issues and rollbacks restore earlier deployments.

Daemons for background workers

Vercel Daemons deploys containers as continuously running background workers on Fluid compute. Users choose the number of instances in each region, and Vercel maintains that capacity by restarting processes when they exit or crash.

Daemons launches with support for pull-based workloads, where the code requests work from another system and processes it as it becomes available. Examples named by Vercel are consuming queue messages or fetching pending jobs from a database.

Vercel illustrates this with a video app. A Dockerfile packages a Python processor, its dependencies and ffmpeg, and the processor prepares a playback copy and thumbnail for each uploaded clip. In the vercel.toml configuration, a Next.js frontend is defined as a service, requests are routed to it, and a daemon named videos points to the Dockerfile. Setting instances.iad1 = 1 keeps one processor instance running in iad1 (Washington, D.C.). In the example, the daemon pulls jobs from Vercel Queues.

What it means for developers

The changes widen the kinds of backend workloads that can run on Vercel. Teams with Python services, custom system libraries or long-running workers can now keep them on the same platform as their frontend instead of splitting them across providers. Vercel says it will keep investing in backend support so more of an existing application can run on the platform, using the languages, dependencies and execution patterns it needs.

Daemons is in Public Beta, so its feature set is still developing. At launch it covers pull-based workloads only.