Claude Doubles 5-Hour Usage Limit via xAI Partnership

Post LinkedIn

🗾Read original on ITmedia AI+ (日本)

#rate-limits #partnerships #computeclaudeclaude anthropic xai

💡Claude doubles 5-hr limit via xAI compute deal—more runtime for your apps!

⚡ 30-Second TL;DR

What Changed

Claude's 5-hour usage limit doubled

Why It Matters

This update benefits heavy Claude users by extending session times, reducing interruptions for developers and researchers. The xAI partnership signals growing compute alliances in AI, potentially stabilizing service reliability amid demand surges.

What To Do Next

Check your Anthropic Console for updated Claude rate limits on your plan.

Who should care:Developers & AI Engineers

Key Points

•Claude's 5-hour usage limit doubled
•Partnership with Elon Musk's SpaceXAI/xAI for compute
•Broader industry collaborations to secure resources
•Partial easing of overall usage restrictions

🧠 Deep Insight

Web-grounded analysis with 11 cited sources.

🔑 Enhanced Key Takeaways

•The partnership grants Anthropic exclusive access to the full 300+ megawatt capacity of the Colossus 1 data center in Memphis, which houses over 220,000 NVIDIA GPUs (including H100, H200, and GB200 accelerators).
•Beyond immediate capacity, the companies have expressed interest in a long-term collaboration to develop multi-gigawatt orbital AI compute infrastructure, citing terrestrial limitations in power, cooling, and land.
•The agreement includes the removal of 'peak hours' usage restrictions for Claude Pro and Max subscribers, alongside significant increases to API rate limits for the Claude Opus model family.

📊 Competitor Analysis▸ Show

Feature	Anthropic (Claude)	OpenAI (ChatGPT)	Google (Gemini)	xAI (Grok)
Primary Compute	AWS, Google, SpaceX (Colossus 1)	Microsoft Azure	Google Cloud	SpaceX (Colossus 2)
Coding Focus	Claude Code (Agentic)	Advanced Data Analysis	Gemini Code Assist	Grok-based coding
Infrastructure Strategy	Multi-cloud + Orbital R&D	Azure-centric	Google-native	Vertical integration (SpaceX)

🛠️ Technical Deep Dive

•Colossus 1 Data Center: Located in Memphis, Tennessee; features 300+ MW power capacity.
•Hardware: 220,000+ NVIDIA GPUs, including H100, H200, and next-generation GB200 accelerators.
•Deployment: Capacity integration for Anthropic systems is scheduled to be fully operational within one month of the May 6, 2026 announcement.
•API Scaling: Tier 1 API limits for Claude Opus models saw increases of up to 1500% for input tokens and 900% for output tokens per minute.

🔮 Future ImplicationsAI analysis grounded in cited sources

SpaceX will emerge as a primary neutral infrastructure provider for AI labs.

By leasing Colossus 1 to a direct competitor (Anthropic), SpaceX is pivoting its AI assets toward a third-party compute-as-a-service business model.

Orbital AI compute will become a serious R&D focus for frontier AI labs by 2030.

The explicit public interest in multi-gigawatt orbital infrastructure signals that terrestrial power and cooling constraints are now considered a long-term existential bottleneck for AI scaling.