According toNewsAnthropic has reached a major cooperation agreement with Elon Musk's SpaceX, and will utilize...xAI's massive computing power located in the Colossus 1 data centerBenefiting from this new injection of computing power, Anthropic announced that it will double the usage cap for Claude Code for paid users and significantly relax other related restrictions.
By establishing a presence in Colossus 1, they are focusing their attention on "orbital computing power".
Under the agreement, Anthropic will be granted access to "all computing capacity" of SpaceX's Colossus 1 data center this month, which is expected to bring Anthropic more than 300 megawatts (MW) of new computing power.
What's even more striking is the ambitious vision for the future shared by both parties. Anthropic explicitly expressed strong interest in further collaboration with SpaceX in its statement, aiming to jointly develop "gigawatt-scale orbital AI computing power." This aligns with SpaceX's recent proposal to the Federal Communications Commission (FCC) to launch one million satellites to establish..."Railway Data Center" The ambitious plan for the Orbital data center undoubtedly opens up a highly futuristic imagination for the future of AI infrastructure.
In addition, Elon Musk also explained that although the computing resources of the Colossus 1 data center are provided to Anthropic, the xAI training computing power source that was recently merged with SpaceX has been transferred to the Colossus 2 data center, so it will not affect the allocation of its computing resources.
Paid users immediately benefit: peak hour limits are unlocked, and API call volume increases significantly.
With ample computing power as a backing, Anthropic immediately returned the resources to frontline users.
For all paid subscribers to the Pro, Max, Team, and Enterprise plans, Anthropic will immediately double the five-hour rate limits for Claude Code. Furthermore, Pro and Max plan users will no longer be bothered by annoying "peak hours" rate throttling; and for the current highest-tier Claude Opus model, the API call limit has also been significantly increased, providing a substantial advantage for development teams that heavily rely on Claude for intensive coding and reasoning.
Elon Musk's attitude takes a dramatic turn.
The reason this alliance has caused such a stir in the industry is largely due to the past grievances between the two parties.
Not long ago, xAI founder Elon Musk fiercely criticized Anthropic as "Misanthropic and evil," and has long held a strong critical attitude towards Anthropic CEO Dario Amodei.
However, Elon Musk's stance on the X platform softened dramatically earlier. He posted that he was "impressed" by the Anthropic team and believed that Claude "should" perform well.
But Elon Musk maintained his characteristic maverick style, adding a strongly personal statement in another post: "If their AI engages in activities that harm humanity, we reserve the right to reclaim that computing power."
Same here.
By way of background for those who care, 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.
Everyone I met was highly competent and cared for a great deal about…
- Elon Musk (@elonmusk) May 6, 2026
