How SpaceXAI & Anthropic Compute Deal Raises Claude Limits

For users and enterprises that cannot get enough of Claude, Anthropic is increasing the usage limits for Claude Code and Claude API. 

This comes as a direct result of the landmark agreement with SpaceXAI, that will give the AI company access to Colossus 1 – one of the world’s largest and fastest-deployed AI supercomputers. 

Anthropic said that the deal will “substantially increase” their compute capacity.

Built in record time, Colossus 1 delivers more than 300 megawatts of compute power through a cluster of over 220,000 NVIDIA GPUs – including H100, H200 and next-generation GB200 accelerators. 

The system is designed to support frontier-scale AI training, inference, multimodal systems and scientific simulations.

“I spent a lot of time last week with senior members of the Anthropic team to understand what they do to ensure Claude is good for humanity and was impressed,” writes Elon Musk, CEO of SpaceXAI, in a post on X. 

“Everyone I met was highly competent and cared a great deal about doing the right thing.

“After that, I was ok leasing Colossus 1 to Anthropic, as SpaceXAI had already moved training to Colossus 2.”

What is the new Claude Usage Limit?

Anthropic has brought significant upgrades to Claude users, this includes “doubling Claude Code’s five-hour rate limits for Pro, Max, Team and seat-based Enterprise plans.”

Peak hour reductions in Claude Code for Pro and Max accounts are now lifted. 

The company is also offering significantly higher rate limits for Claude Opus models. 

In addition to the SpaceX deal, Anthropic’s recent 5GW agreement with Amazon, similar capacity deal with Google and Broadcom, along with partnership with Microsoft and NVIDIA that brings in US$30bn of Azure capacity, all contributed to these increased Claude usage limits. 

Orbital AI Compute Moves Closer to Reality

The new partnership extends beyond the restraints of terrestrial infrastructure as Anthropic also expressed interest in partnering with SpaceXAI to develop multiple gigawatts of orbital AI compute capacity. 

In an AI driven world, compute is the most valuable commodity. With the energy needs of training and operating next-generation AI systems increasing faster than traditional power, cooling and land resources can catch up, technology firms are exploring alternative approaches.

“SpaceX is the only organisation with the launch cadence, mass-to-orbit economics and constellation operations experience to make orbital compute a near-term engineering program rather than a research concept,” the SpaceXAI press release reads. 

“If engineering challenges can be overcome, space-based compute offers near-limitless sustainable power with less impact on Earth.”

The partnership with Anthropic signals that the future of AI may depend as much on infrastructure innovation as advances in the models themselves.

AI Companies Race to Secure Frontier-Scale Compute

Earlier this year, Cursor announced it was partnering with SpaceXAI to accelerate training for its Composer coding models using the same Colossus infrastructure.

In its announcement, Cursor said each increase in compute had translated into “meaningfully more capable models”, adding that the company had been “bottlenecked by compute” before securing access to the platform. 

This is reflective of the underlying fact that in the AI sector infrastructure availability is preliminary in determining how quickly companies can develop new and better systems.

As developers and enterprise users are adopting AI products at scale never seen before, demand for reliable compute capacity has skyrocketed. 

As Anthropic CEO Dario Amodei reflects that his company has tried to plan for a 10-fold growth, meanwhile the revenue and usage increased 80-fold in the first quarter – when taken in an annualised basis. 

“That is the reason we have had difficulties with compute,” Amodei says, adding that the company is “working as quickly as possible to provide more”.