Back

Anthropic Doubles Claude Code Limits After Securing SpaceX Colossus 1 Capacity

Anthropic rolled out three immediate changes to its Claude AI services on Tuesday, signaling a major upgrade in capacity for both its chat‑based Claude Code and the Claude Opus API. The most visible shift is a doubling of the five‑hour usage windows for Claude Code on the Pro, Max, Team and seat‑based Enterprise plans. In addition, the company eliminated the peak‑hour throttling that previously limited Pro and Max accounts during high‑traffic periods. The third adjustment raises the ceiling on API calls for Claude Opus models, with the new limits published alongside the announcement.

Behind the upgrade lies a fresh compute agreement with SpaceX. Anthropic has signed to take the entire compute capacity of SpaceX’s Colossus 1 data centre, a facility that will deliver more than 300 megawatts of power and house upwards of 220,000 Nvidia GPUs within the month. The influx of hardware directly fuels the higher rate limits, allowing the firm to serve Claude Pro and Claude Max subscribers with less congestion and faster response times.

SpaceX partnership expands Anthropic’s hardware pool

The Colossus 1 deal joins a series of sizable compute commitments Anthropic has disclosed in recent months. The company already holds a up‑to‑5‑gigawatt arrangement with Amazon, slated to bring nearly 1 GW of capacity online by the end of 2026, a 5‑GW agreement with Google and Broadcom expected to start delivering power in 2027, and a strategic partnership with Microsoft and Nvidia that includes $30 billion of Azure capacity. A $50 billion investment with Fluidstack further underscores Anthropic’s aggressive push to secure AI‑specific infrastructure.

Anthropic’s hardware mix spans AWS Trainium chips, Google TPUs and Nvidia GPUs, but the new SpaceX resources represent a significant addition to that portfolio. While the firm has expressed interest in collaborating with SpaceX on orbital AI compute—a venture that could eventually deliver multiple gigawatts of processing power in space—no formal agreement on that front has been signed yet.

Beyond raw compute, Anthropic is tailoring its expansion to meet regulatory and compliance demands. Recent collaborations with Amazon will add inference capacity in Asia and Europe, addressing enterprise customers in regulated sectors that require data residency and local processing. The company says it will continue to partner only with “democratic” nations whose legal frameworks can support large‑scale AI investments.

Anthropic also reaffirmed an earlier pledge to offset any consumer electricity‑price hikes caused by its U.S. data centres, and hinted at extending that commitment as it moves into new jurisdictions. The company’s focus on deliberate, region‑specific growth aims to balance performance gains with responsible stewardship of energy resources.

For users, the headline changes are clear. Pro and Max subscribers now enjoy a doubled five‑hour usage window and no longer face throttling during peak periods. The Opus API limits have likewise risen, giving developers more leeway to integrate Claude’s capabilities into their applications. Anthropic’s statement positions the upgrades as a direct benefit of the SpaceX partnership, promising a smoother, faster experience for its growing customer base.

Used: News Factory APP - news discovery and automation - ChatGPT for Business

Source: The Next Web

Also available in: