Anthropic's Claude Code Tops $2.5B ARR, Finds Crypto Miner

Anthropic’s Claude Code, the company’s AI-powered coding assistant, is experiencing rapid growth with annualized revenue exceeding $2.5 billion as of February 2026, according to recent reports. The product has become Anthropic’s fastest-growing offering, with enterprise customers accounting for more than half of the coding tool’s revenue. This commercial success comes as the company continues to roll out new capabilities and bug fixes across recent software releases.

The coding assistant has attracted attention not only for its revenue performance but also for its practical applications. Claude Code reportedly detected an XMRig cryptocurrency miner consuming roughly 1,175% CPU and driving a developer’s Mac to 112°C. The developer subsequently used the AI agent to help remove the malicious software, demonstrating the tool’s utility beyond standard coding tasks.

Recent Product Updates and Capabilities

Claude Code version 2.1.302 made fork subagents the default across Claude Code’s command-line interface, Agent SDK, desktop app, and VS Code integration. This update represents a significant change in how the tool handles concurrent operations and workflow management.

The 2.1.302 release allows subagents to run in the background in Claude’s desktop app and VS Code. This background execution capability enables developers to continue working while the AI handles time-consuming tasks. Version 2.1.303 subsequently focused on bug fixes and reliability improvements.

Claude Code reportedly added automatic compaction for long terminal conversations after periods of inactivity, helping preserve context while limiting costs. This feature addresses one of the ongoing challenges in AI-assisted development: managing conversation length and associated token costs while maintaining sufficient context for the AI to provide useful assistance.

Security and Permission Updates

Recent releases have addressed several security-related issues. The 2.1.302 release also changed default permission behavior and included plugin and filesystem fixes. The update also tightened MCP permission rules and aligned subagent defaults across supported clients.

Recent Claude Code releases fixed a regression that could cause cloud and headless sessions to execute an approved tool call twice. This bug had the potential to create unintended consequences when developers approved specific actions. Another fix addressed instruction-based hooks that could fail to block commands as intended, representing an important security improvement.

Enterprise Features Expansion

Anthropic reportedly moved its managed Code Review service out of research preview for Team and Enterprise organizations using GitHub. The general availability of this feature marks a shift from experimental to production-ready status for business customers.

The service is billed by token usage rather than a fixed price per review, aligning with the variable pricing model common across AI service offerings. This approach allows organizations to pay based on actual usage rather than committing to fixed fees regardless of review volume or complexity.

Key Facts

Claude Code’s annualized revenue exceeded $2.5 billion by February 2026
Enterprise customers account for more than half of the coding tool’s revenue
Version 2.1.302 made fork subagents the default across all Claude Code clients
Code Review service moved to general availability for Team and Enterprise GitHub users
Claude Code detected a cryptocurrency miner consuming roughly 1,175% CPU on a developer’s Mac

Market Context

The strong revenue performance positions Claude Code as a major player in the AI-assisted development tools market. The product competes in an increasingly crowded space where multiple vendors offer AI coding assistants, code completion, and automated review capabilities.

The emphasis on enterprise customers reflects a broader industry trend toward business-focused AI tool adoption. Organizations are integrating AI assistants into development workflows as they seek to improve productivity and code quality while managing development costs.

The cryptocurrency miner detection incident highlights an unexpected use case for AI coding assistants. While these tools are primarily designed to help write and review code, their ability to analyze system behavior and identify anomalies demonstrates potential applications in security and system administration contexts.

Sources

Written by

Mike Gingerich

Author

Author: Mike Gingerich, President of web firm Digital Hill, Co-Founder of TabSite .
Digital and Social Media Marketer, Speaker, and Business Consultant. Part geek, part marketer, total digital junkie! Seeking to add value, make the complex simple, and leave a positive impact.

Follow me on twitter: @mike_gingerich.