Anthropic Secures SpaceX Compute Partnership, Doubles Claude Code Limits and Boosts Opus API Rate Limits
Patch your Claude Code integration to remove peak‑hour limit checks and double your 5‑hour quota handling, and update Opus API calls to accommodate higher rate limits.
Patch your Claude Code integration to remove peak‑hour limit checks and double your 5‑hour quota handling, and update Opus API calls to accommodate higher rate limits.
Summary
Anthropic announced a compute partnership with SpaceX that will give it access to the Colossus 1 supercomputer, effectively adding more than 300 MW of new capacity and roughly 220,000 GPUs, including 150,000 H100s, 50,000 H200s, and 30,000 GB200s, and enabling inference to ramp on Colossus within days.
The partnership immediately translates into higher usage limits for Claude products: the 5‑hour rate limits for Pro, Max, Team, and seat‑based Enterprise are doubled; peak‑hour limit reductions are removed for Pro and Max; and Opus API rate limits are substantially increased, while weekly limits remain unchanged.
Anthropic clarified that the compute boost is a short‑term expansion, with the inference ramp expected within days, and that the increased limits will help users who hit 5‑hour caps but not yet those who hit weekly caps, with further changes likely as compute lands.
Key changes
- SpaceX compute partnership gives Anthropic access to Colossus 1 with >300 MW capacity and ~220,000 GPUs (150k H100s, 50k H200s, 30k GB200s)
- Claude Code 5‑hour rate limits for Pro, Max, Team, and seat‑based Enterprise are doubled immediately
- Peak‑hour limit reductions for Pro and Max are removed
- Opus API rate limits are substantially increased
- Weekly limits remain unchanged
- Inference on Colossus will ramp within days
- Compute partnership expected to increase capacity for Claude inference