Python Workers Are Now Generally Available on Cloudflare

Cloudflare announced on September 21, 2026, that Python Workers are now generally available — marking Python as a first-class, fully supported language on the Cloudflare Developer Platform alongside TypeScript and JavaScript.

For Python developers, this means you can deploy applications using the frameworks, libraries, and patterns you already know, connected natively to Cloudflare's edge infrastructure including Workers AI, R2 storage, D1 databases, Hyperdrive, Durable Objects, Queues, and Workflows.

Two Years in the Making

Cloudflare introduced Python Workers in 2024 with an ambitious goal: make writing Workers in Python as straightforward as writing them in TypeScript, and make the Python ecosystem "just work" on the edge runtime.

That goal required solving hard technical problems. Python's dynamic nature, extensive standard library, and C extension dependencies do not map cleanly to V8 isolates — the lightweight sandbox Cloudflare uses for JavaScript Workers. The engineering team built a compatibility layer that handles package resolution, native module support, and runtime differences transparently.

General availability means the team is confident in stability, performance, and ecosystem coverage for production workloads.

What You Can Build

Python Workers GA unlocks several categories of applications that were previously impractical on Cloudflare's platform:

Web Frameworks

FastAPI, Django, and Flask all run natively in Python Workers. Cloudflare implemented a built-in connector that routes HTTP requests to your web application without requiring custom adapter code. If you have an existing FastAPI application, the migration path involves packaging it as a Worker rather than rewriting it for a different runtime.

AI and Machine Learning

Python's dominance in AI makes this the most significant capability. Libraries like openai, langchain, and mcp (Model Context Protocol) run natively in Python Workers. You can combine them with Workers AI for serverless GPU inference on Cloudflare's network, or proxy requests through Cloudflare AI Gateway for rate limiting, caching, and observability.

A practical example Cloudflare demonstrated: an AI-driven image-to-image generator that accepts user requests, queues them via Cloudflare Queues, orchestrates generation through Workflows, runs inference on Workers AI, and stores results in an R2 bucket. The entire pipeline runs in Python on Cloudflare's edge.

Database Access

Python Workers integrate with Hyperdrive for relational database connections. If your application uses PostgreSQL or MySQL, Hyperdrive provides connection pooling and query caching at the edge — reducing latency for database-backed Python APIs deployed globally.

D1, Cloudflare's SQLite-based edge database, is also accessible from Python Workers for applications that need lightweight, distributed data storage without managing a separate database server.

Background Processing

Cloudflare Queues and Workflows enable asynchronous processing patterns. A Python Worker can accept an HTTP request, enqueue a job, and return immediately while a separate Worker processes the job. Workflows add orchestration for multi-step pipelines with failure handling and retries.

Dynamic Workers: Python Inside Python

One of the more novel capabilities is Dynamic Workers — the ability to create a Python Worker inside another Worker at runtime. This enables patterns like user-submitted code execution, plugin systems, and multi-tenant architectures where each tenant runs isolated Python code within a shared Worker.

Migration Guide for Python Developers

If you are considering moving a Python application to Cloudflare Workers, here is a practical assessment framework:

Good fit:

• API servers built with FastAPI, Flask, or Django
• AI/ML inference pipelines using standard Python libraries
• Background job processors that benefit from global edge deployment
• Applications with moderate memory and CPU requirements

Requires adaptation:

• Applications with heavy C extension dependencies not yet supported
• Long-running processes that exceed Worker execution time limits
• Applications requiring persistent local filesystem access
• Workloads with very high memory requirements

Start here:

1. Install the Cloudflare Workers CLI (wrangler)
2. Create a new Python Worker project: wrangler init --lang python
3. Add your framework (e.g., fastapi) to requirements.txt
4. Connect to Cloudflare services (R2, D1, Hyperdrive) via bindings in wrangler.toml
5. Deploy: wrangler deploy

Performance and Cold Starts

Edge runtimes always involve tradeoffs around cold starts and execution limits. Cloudflare's V8 isolate model minimizes cold start latency compared to container-based serverless platforms, but Python Workers may have slightly higher initialization times than JavaScript Workers due to Python runtime setup.

For latency-sensitive applications, test cold start behavior in your target regions. Cloudflare's global network of data centers means most users connect to a nearby edge location, but the first request to a cold Worker still incurs initialization overhead.

The Bigger Picture

Python Workers GA is part of Cloudflare's broader strategy to become the default platform for AI-native application development. The Agent Development Lifecycle, Workflows orchestration, agent observability, and now full Python support create an integrated stack for building, deploying, and monitoring AI-powered applications at the edge.

For Python developers who have watched the serverless ecosystem evolve primarily around JavaScript and Go, this is a meaningful shift. The language that powers most AI research and data science work can now run on one of the world's largest edge networks — with direct access to GPU inference, object storage, and distributed databases.

The question is no longer whether you can run Python at the edge. It is what you will build now that you can.