SAN FRANCISCO, January 13, 2026 — Anthropic has released Claude Code, a new AI coding tool integrated into its Claude chatbot, designed to help developers write, debug, and understand code more efficiently.
Available now in beta to all Claude Pro, Team, and Enterprise users, Claude Code allows the model to read code files, suggest edits, explain complex logic, generate tests, and perform multi-step programming tasks directly in the conversation interface. It supports major languages, including Python , JavaScript, TypeScript, Java, C++, Go, and Rust.
Anthropic emphasized the tool’s focus on safety and reliability. Claude Code includes safeguards against generating malicious code and provides transparent reasoning for suggestions. Users can ask the model to “think step by step” or “show your work” to see the logic behind recommendations.
The company highlighted early user feedback showing improved productivity, particularly for debugging and refactoring. In internal tests, Claude Code resolved real GitHub issues faster than previous versions of the model. It builds on Claude 3.5 Sonnet’s strong coding performance, which has topped several independent benchmarks.
Unlike some competitors that require separate IDE plugins or code interpreters, Claude Code operates natively within the existing Claude web and mobile apps. Users can upload code files up to 200,000 tokens (roughly 150,000 lines) for analysis.
Anthropic plans to expand the beta based on feedback, with potential additions including better multi-file support, version control integration, and real-time collaboration features.
