Replicate is Joining Cloudflare

This post is also available in Traditional Chinese, German, Japanese, Korean, Spanish (LatAm), Dutch, and French.

Today marks a significant announcement: Replicate, the leading platform for running AI models, is joining Cloudflare.

Our initial conversations with Replicate revealed a deep alignment beyond just a shared appreciation for vibrant color palettes. Cloudflare's Workers developer platform has always aimed to simplify the building and deployment of full-stack applications. Similarly, Replicate has been dedicated to making AI model deployment as straightforward as writing a single line of code. We recognized an incredible opportunity to create something even more powerful by integrating the Replicate platform directly into Cloudflare.

We are thrilled to share this news and are even more excited about the benefits it will bring to our customers. Incorporating Replicate's robust tools into Cloudflare's Developer Platform will further establish it as the premier destination for building and deploying any AI or agentic workflow on the Internet.

What Does This Mean For You?

Before delving into the future of AI, we want to address the most pressing questions for current Replicate and Cloudflare users. In summary:

• For existing Replicate users: Your APIs and workflows will continue to operate without interruption. You will soon experience enhanced performance and reliability thanks to Cloudflare's global network.
• For existing Workers AI users: Prepare for a massive expansion of our model catalog and the new capability to run fine-tuned and custom models directly on Workers AI.

Now, let's explore why we're so enthusiastic about our shared future.

The AI Revolution: Fueled by Open Source

Long before "AI" became a ubiquitous term, captivating every conversation, it was known for decades as "machine learning"—a specialized, almost academic field. Progress was steady but often isolated, with breakthroughs primarily confined to a few large, well-funded research labs. Models were monolithic, data proprietary, and tools largely inaccessible to most developers.

Everything changed when the spirit of open-source collaboration—the very force that built the modern Internet—converged with machine learning. Researchers and companies began publishing not just their papers, but their model weights and code.

This ignited an extraordinary explosion of innovation. The pace of change in the last few years has been staggering; what was considered state-of-the-art 18 months ago (or, at times, just days ago) is now the baseline. This acceleration is most evident in generative AI.

We witnessed a rapid evolution from uncanny, blurry curiosities to photorealistic image generation in what felt like the blink of an eye. Open-source models such as Stable Diffusion immediately unleashed creativity for developers, and that was just the beginning. Today, Replicate's model catalog showcases thousands of image models across nearly every variation, each building upon the last.

This progression wasn't limited to image models; it extended to video, audio, language models, and more.

However, this incredible, community-driven progress created a significant practical challenge: How do you actually run these models? Each new model comes with different dependencies, requires specific (and often substantial) GPU hardware, and needs complex serving infrastructure to scale effectively. Developers often found themselves spending more time battling CUDA drivers and requirements.txt files than actually developing their applications.

This is precisely the problem Replicate solved. They built a platform that abstracts away all this complexity, utilizing their open-source tool Cog to package models into standard, reproducible containers. This allows any developer or data scientist to run even the most complex open-source models with a simple API call.

Today, Replicate's catalog boasts over 50,000 open-source and fine-tuned models. While open source unlocked numerous possibilities, Replicate's toolset goes further, enabling developers to access any models they need from a single location. Their marketplace also provides seamless access to leading proprietary models like GPT-5 and Claude Sonnet, all through the same unified API.

Crucially, Replicate didn't just build an inference service; they cultivated a community. Much innovation springs from being inspired by others' work, iterating on it, and improving it. Replicate has become the definitive hub for developers to discover, share, fine-tune, and experiment with the latest models in a public playground.

Stronger Together: The AI Catalog Meets the AI Cloud

Returning to the Workers Platform mission: our continuous goal has been to empower developers to build full-stack applications without the burden of infrastructure management. While this core mission remains unchanged, AI has profoundly altered the requirements for modern applications.

The types of applications developers are creating are evolving rapidly. Three years ago, agents or AI-generated launch videos were unheard of; today, they are commonplace. Consequently, what developers need and expect from the cloud—or the AI cloud—has also transformed.

To meet these evolving developer needs, Cloudflare has been meticulously constructing the foundational pillars of its AI Cloud, designed to execute inference at the edge, closer to users. This isn't just a single product but an entire integrated stack:

• Workers AI: Serverless GPU inference delivered on our global network.
• AI Gateway: A control plane offering caching, rate-limiting, and observability for any AI API.
• Data Stack: Including Vectorize (our vector database) and R2 (for robust model and data storage).
• Orchestration: Tools such as AI Search (formerly Autorag), Agents, and Workflows to construct complex, multi-step applications.
• Foundation: All built upon our core developer platform, encompassing Workers, Durable Objects, and the rest of our robust stack.

While we've been helping developers scale their applications, Replicate has pursued a parallel mission: making AI model deployment as simple as deploying code. This is where our synergies converge. Replicate brings one of the industry's largest and most vibrant model catalogs and developer communities. Cloudflare contributes an incredibly performant global network and serverless inference platform. Together, we can offer the best of both worlds: the most comprehensive selection of models, runnable on a fast, reliable, and affordable inference platform.

Our Shared Vision

For the Community: The Hub for AI Exploration

The ability to share models, publish fine-tunes, earn stars, and experiment in the playground is central to the Replicate community experience. We are committed to continued investment in and growth of this platform as the premier destination for AI discovery and experimentation, now supercharged by Cloudflare's global network for an even faster, more responsive experience for everyone.

The Future of Inference: One Platform, All Models

Our vision is to merge the best features of both platforms. We will integrate the entire Replicate catalog—all 50,000+ models and fine-tunes—into Workers AI. This provides ultimate flexibility: run models in Replicate's adaptable environment or on Cloudflare's serverless platform, all from a unified interface.

But our expansion goes beyond the catalog. We are excited to announce the introduction of fine-tuning capabilities to Workers AI, powered by Replicate's deep expertise. Furthermore, Workers AI will become more flexible than ever. Soon, you'll be able to bring your own custom models to our network. We will leverage Replicate's proficiency with Cog to ensure this process is seamless, reproducible, and easy.

The AI Cloud: More Than Just Inference

Running a model is merely one piece of the puzzle. The true innovation emerges when you connect AI to your entire application ecosystem. Imagine the possibilities when Replicate's massive catalog is deeply integrated with the entire Cloudflare developer platform: run a model and store results directly in R2 or Vectorize; trigger inference from a Worker or Queue; utilize Durable Objects to manage state for an AI agent; or build real-time generative UIs with WebRTC and WebSockets.

To manage this intricate environment, we will integrate our unified inference platform deeply with the AI Gateway, providing a single control plane for observability, prompt management, A/B testing, and cost analytics across all your models, whether they are running on Cloudflare, Replicate, or any other provider.

Welcome to the Team!

We are incredibly excited to welcome the Replicate team to Cloudflare. Their passion for the developer community and their unparalleled expertise in the AI ecosystem are invaluable. We eagerly anticipate building the future of AI together.

Cloudflare's connectivity cloud secures entire corporate networks, enables customers to build Internet-scale applications efficiently, accelerates any website or Internet application, wards off DDoS attacks, keeps hackers at bay, and assists you on your journey to Zero Trust.

Visit 1.1.1.1 from any device to get started with our free app that makes your Internet faster and safer.

To learn more about our mission to help build a better Internet, start here. If you're looking for a new career direction, check out our open positions.