Anthropic doubles Claude Code paid-use allowance from 5 to 10 hours

Anthropic just gave Claude Code users more room to work — and the change is pretty concrete. On May 6, the company said it doubled Claude Code’s five-hour rate limits for Pro, Max, Team, and seat-based Enterprise plans. It also removed the peak-hours limit reduction for Pro and Max accounts. At the same time, Anthropic said it had signed a compute deal with SpaceX that should add a lot more capacity fast. ### What actually changed for users? The headline change is simple: the old five-hour Claude Code allowance for paid plans is now twice as large. Anthropic framed it as effective immediately on May 6. For individual subscribers, Pro and Max also stop getting penalized during peak hours, which matters because those slowdowns made usage feel unpredictable even when you were paying for access. Anthropic also said it raised API rate limits for Claude Opus models, so this was not just a chat-app tweak — it was a broader capacity release. (anthropic.com) ### Why was Claude Code hitting limits so often? Claude Code is not a lightweight autocomplete tool. It reads repositories, edits files, runs tests, and can chew through a lot of context in one session. Anthropic’s own product page pitches it as an agentic coding system, and the help docs make clear that Claude Code shares usage budgets with Claude across web, desktop, mobile, and terminal access on Pro and Max. (anthropic.com) So a heavy coding session and a heavy chat session can eat from the same pool. That is why users felt the cap as a real workflow problem, not an abstract quota. ### Where is the extra capacity coming from? Basically, Anthropic bought more compute. The company said it signed an agreement to use all of the compute capacity at SpaceX’s Colossus 1 data center. Anthropic put a huge number on that — more than 300 megawatts of new capacity, or over 220,000 Nvidia GPUs, within the month. It explicitly said that extra capacity would directly improve service for Claude Pro and Claude Max subscribers. (anthropic.com) ### Why does the SpaceX angle matter? Because this is the real bottleneck in AI right now. Better models help, but if you do not have enough chips and power, users slam into rate limits. Anthropic’s post basically said the quiet part out loud: more compute deals are what made higher Claude Code and API limits possible. The SpaceX agreement also sits beside other giant infrastructure deals Anthropic listed with Amazon, Google, Broadcom, Microsoft, Nvidia, and Fluidstack. (anthropic.com) ### Who gets this on business plans? Team customers get Claude Code included with every seat, and premium seats get more usage. Enterprise is a little messier. On new and self-serve Enterprise plans, Claude Code is included with the Enterprise seat. On older Enterprise setups, access depends on seat type — like Chat + Claude Code seats or Premium seats. Team and seat-based Enterprise plans can also buy extra usage so members can keep working after they hit included limits. (anthropic.com) ### What is the catch for regulated customers? The catch is HIPAA-ready Enterprise. Anthropic’s help center says Claude Code can be bundled into those seats, but Claude Code usage itself is not covered under the HIPAA-ready offering. Only the chat functionality is covered in that setup. Separately, Anthropic’s BAA docs say beta features like Cowork are not covered either. So healthcare and other regulated buyers do get more clarity here, but not full parity across every product surface. (support.claude.com) ### Why does this matter beyond one product tweak? Claude Code is one of Anthropic’s strongest hooks with developers and companies, and usage friction is deadly for tools people want to keep open all day. Doubling the five-hour window does not mean usage is unlimited — weekly caps and plan structure still matter — but it makes the product feel more like infrastructure and less like a rationed demo. That is the bigger story. (support.claude.com) Anthropic is spending hard on compute so it can sell reliability, not just model quality. ### Bottom line? This was not a vague “we improved limits” announcement. Anthropic tied a very specific user-facing upgrade to a very specific compute land grab. If you use Claude Code heavily, the practical change is simple — longer uninterrupted sessions now, and fewer peak-time headaches if you are on Pro or Max. (anthropic.com)