Anthropic signs with SpaceX for 220,000+ NVIDIA GPUs at Colossus 1, doubles Claude Code rate limits, and removes peak-hour throttling.
Key Takeaways
SpaceX Colossus 1 deal adds 300 MW of capacity; Anthropic gains access within the month for Pro and Max subscribers.
Claude Code five-hour rate limits double for Pro, Max, Team, and seat-based Enterprise plans, effective immediately.
Peak-hours limit reduction on Claude Code removed for Pro and Max accounts.
API rate limits for Claude Opus models raised considerably; specific tier table published in the post.
Total announced compute pipeline: Amazon (up to 5 GW), Google/Broadcom (5 GW), Microsoft/NVIDIA ($30B Azure), Fluidstack ($50B), plus orbital compute interest with SpaceX.
Hacker News Comment Review
Commenters broadly noted the five-hour doubling is hollow without a matching weekly limit increase; users who hit hourly caps benefit, but total weekly throughput is unchanged for heavy users.
The orbital compute mention drew skepticism: consensus leans toward it being a negotiating concession to SpaceX rather than a serious near-term technical plan.
Environmental concerns surfaced around Colossus 1’s Memphis power situation, with debate over whether “illegal” power usage is accurate or overstated.
Notable Comments
@arian_: “Anthropic renting out the data center Elon built for Grok is the kind of plot twist you can’t make up.”
@brokencode: SpaceX converts a Grok cost center into major revenue ahead of its IPO – sharp read on the business angle.
