Anthropic is increasing usage limits for Claude after signing a compute partnership with SpaceX, giving the company more capacity to support heavier Claude usage across coding, API, and paid subscription plans.

In its official announcement, Anthropic said the immediate changes include doubled five-hour limits for Claude Code, higher API rate limits for Claude Opus models, and the removal of peak-hour reductions for Claude Code on Pro and Max accounts. The company said the updates are aimed at improving the experience for its “most dedicated customers.”

Anthropic is easing the pressure around Claude usage

The announcement lands at a time when Claude’s limits have become a larger part of the product conversation.

As more users started relying on Claude for longer tasks, usage limits became more visible. This is especially true for Claude Code and API users, where sessions can involve longer context, repeated requests, larger outputs, and more compute-heavy workflows than a normal chat.

That is why this update is not just about giving users more messages. It is about Anthropic trying to make Claude feel less constrained for people who use it as part of their actual work.

The biggest immediate gains are for Claude Code and Opus API users

For Claude Code, Anthropic is doubling five-hour rate limits for Pro, Max, Team, and seat-based Enterprise plans. Pro and Max users will also no longer face reduced Claude Code limits during peak hours.

The API side is just as important. Anthropic said it is raising rate limits “considerably” for Claude Opus models, which matters for developers and companies building heavier workflows on top of Claude.

So the update has two faces. Claude Code users get more room inside the product, while API users get more throughput for larger or more frequent requests.

SpaceX is the capacity behind the limit increase

The reason Anthropic can make these changes is the new SpaceX compute deal.

Anthropic said it has signed an agreement to use all compute capacity at SpaceX’s Colossus 1 data center. That gives the company access to more than 300 megawatts of new capacity, equal to over 220,000 Nvidia GPUs, within the month.

The company also said this extra capacity will directly improve capacity for Claude Pro and Claude Max subscribers. That line matters because it shows the SpaceX deal is not only about Claude Code. It is part of Anthropic’s broader effort to make Claude’s paid plans easier to scale as demand grows.

This is part of a much larger compute race

Anthropic also used the announcement to point to its wider infrastructure buildout.

The company mentioned recent compute agreements with Amazon, Google and Broadcom, Microsoft and Nvidia, and Fluidstack. These deals include future capacity across AWS Trainium, Google TPUs, Nvidia GPUs, and new AI infrastructure investments.

That detail gives the SpaceX deal more context. Anthropic is not treating higher Claude limits as a small product adjustment. It is trying to secure enough compute to keep Claude useful as more people use it for coding, agents, enterprise workflows, and API-based applications.

Higher limits do not mean unlimited Claude

Claude is not becoming unlimited, and users can still hit caps when they push the product heavily.

But the direction is important. After a period where limits made Claude feel tighter for many power users, Anthropic is now using new compute capacity to loosen some of those constraints.

For everyday users, the most practical takeaway is simple: Claude should have more breathing room, especially for paid Claude Code users, Pro and Max subscribers, and developers building with Opus.

The real test is whether that added capacity makes Claude feel more dependable during serious work, not just more generous on paper.