Anthropic Scales with SpaceX: Claude Code Limits Doubled, 220K+ GPUs Added Within a Month
Anthropic raised Claude limits and signed a SpaceX compute deal: Claude Code 5-hour caps doubled, Pro/Max peak cuts removed, Opus API limits raised, plus 300+ MW and 220K+ GPUs.
Published 146 days ago. Content may be outdated.
On May 6, 2026, Anthropic announced two major updates at once:
- Higher usage limits for Claude (effective immediately)
- A new compute agreement with SpaceX
Official source: Higher usage limits for Claude and a compute deal with SpaceX
30-Second Summary
- Claude Code 5-hour limits are doubled (Pro, Max, Team, and seat-based Enterprise)
- Peak-hour limit reductions are removed for Pro and Max on Claude Code
- Claude Opus API rate limits are raised considerably
- Anthropic will use all compute capacity at SpaceX Colossus 1
- 300+ MW of new capacity (220,000+ NVIDIA GPUs) within one month
If you rely on long coding sessions, agentic workflows, or high-throughput API usage, this update is directly relevant.
What Changed, and Who Benefits?
| Change | Applies To | Practical Impact |
|---|---|---|
| Claude Code 5-hour limits doubled | Pro / Max / Team / seat-based Enterprise | Fewer interruptions during long sessions |
| Peak-hour reduction removed | Pro / Max | More stable performance during busy periods |
| Higher Opus API rate limits | API developers and enterprises | Higher concurrency and throughput ceilings |
Bottom line: this is not just “more quota”. It is a reliability move for heavy usage patterns.
What the SpaceX Deal Adds
Anthropic states it has signed an agreement to use all compute capacity at SpaceX’s Colossus 1 datacenter, with:
- 300+ MW of additional capacity
- 220,000+ NVIDIA GPUs
- Availability targeted within one month
Anthropic also says this added capacity will directly improve service capacity for Claude Pro and Claude Max subscribers.
Part of a Larger Expansion Program
The SpaceX agreement sits alongside other recent compute commitments noted by Anthropic:
- Up to 5 GW with Amazon (including nearly 1 GW by end of 2026)
- 5 GW with Google and Broadcom (starting in 2027)
- Strategic Microsoft and NVIDIA partnership (including $30B of Azure capacity)
- $50B U.S. AI infrastructure investment with Fluidstack
This signals a broader strategy: scaling model capability and delivery capacity in parallel.
What It Means for Teams
- Individual developers: longer uninterrupted Claude Code sessions
- Teams: less peak-hour variability in day-to-day collaboration
- Enterprise API users: better headroom for production-grade Opus workloads
Anthropic also reiterated international expansion plans to support compliance and data residency needs in regulated sectors.
Closing Takeaway
The message from this update is clear: frontier model competition is now about intelligence and dependable large-scale delivery, not just benchmark quality.
More Articles
