We have formed a partnership with SpaceX that will significantly expand our computing resources. This, combined with our other recent compute agreements, has allowed us to raise the usage limits for both Claude Code and the Claude API. Below, we outline these updates and the strides we’re making on compute capacity.
Higher usage limits. Here is the paraphrased version:
We’re implementing three improvements today to enhance the Claude experience for our most loyal users. First, we’ve doubled the five-hour rate limits for Claude Code across Pro, Max, Team, and seat-based Enterprise plans. Second, we’ve eliminated the peak-hour rate limit reductions for Claude Code on Pro and Max accounts. Third, we’re significantly increasing our API rate limits for Claude Opus models, as detailed in the table below. Here’s the paraphrased version:
We’ve refreshed the API rate limits for our Claude Opus models. We’ve also formed a new compute partnership with SpaceX. We have entered into an agreement with SpaceX to utilize the entire compute capacity of their Colossus 1 data center. This provides us with access to over 300 megawatts of additional capacity (more than 220,000 NVIDIA GPUs) within the month. This extra capacity will directly benefit Claude Pro and Claude Max subscribers. It adds to our other major compute announcements. An agreement with Amazon for up to 5 GW, including almost 1 GW of additional capacity by the end of 2026.
Anthropic News