Anthropic Raises Claude Code Usage Limits, Credits SpaceX Deal

Anthropic has increased usage limits for Claude Code, its terminal-based agentic coding tool, citing expanded infrastructure capacity enabled in part by a new enterprise agreement with SpaceX. The move signals both growing demand for AI-assisted software development and Anthropic's effort to keep pace with that demand operationally.

Claude Code was released in general availability earlier this year and quickly became one of the more capable tools in the agentic coding category. Its ability to operate across full codebases — reading files, running commands, and executing multi-step development tasks — made it compute-intensive by design. Usage limits had been a friction point for engineers and teams pushing it hard in production workflows.

The SpaceX deal represents a notable pairing: a frontier AI safety company providing services to one of the most technically demanding private aerospace operators in the world. While financial details have not been disclosed, the implication is that enterprise-scale contracts are now materially influencing Anthropic's capacity planning decisions.

The core change is straightforward — developers using Claude Code will encounter rate limits less frequently, allowing for longer uninterrupted sessions and more intensive agentic tasks. For teams running Claude Code as a continuous development assistant rather than a one-off query tool, this is a practical improvement. The prior limits created workflow interruptions that undermined the value proposition of an agent meant to operate autonomously across extended tasks.

The SpaceX relationship is worth examining beyond the headline. SpaceX operates a highly complex internal engineering environment — across avionics, propulsion, software systems, and ground infrastructure. Using an external AI coding agent at that level of technical sophistication suggests the organization is either running Claude Code for bounded, lower-risk development tasks, or has worked out sufficient security and access controls to allow deeper integration. Either way, it represents an enterprise validation that carries weight.

For Anthropic, the ability to point to a high-profile enterprise customer as a direct contributor to infrastructure expansion is a useful signal to the market. It frames Claude Code not as a developer productivity toy but as a serious enterprise tool with capacity requirements that scale alongside real deployment.

The broader implication for the agentic coding market is that compute capacity is increasingly a competitive variable. As developers move from using AI assistants in short conversational exchanges to running continuous agents against large codebases, the infrastructure requirements are categorically different. Rate limits are not just a user experience issue — they define whether a tool is viable for production use. Anthropic is treating this as a priority.

This also reflects a maturing enterprise go-to-market for Anthropic overall. The company has historically led with research credibility and model capability. Larger deals with engineering-intensive organizations suggest it is now building the sales infrastructure and capacity agreements to serve enterprise workloads at scale — a necessary evolution if Claude is to compete with OpenAI and Google in the enterprise AI segment where volume and reliability matter as much as raw capability.

The convergence of limit increases, enterprise validation, and capacity expansion tied to real customer contracts suggests Anthropic is moving Claude Code from a capable but constrained tool to one it is prepared to back operationally at production scale.

Sources: — Ars Technica (https://arstechnica.com/ai/2026/05/anthropic-raises-claude-code-usage-limits-credits-new-deal-with-spacex/)