Anthropic's agentic coding tool — Claude in your terminal, IDE, desktop app and browser, reading and editing your codebase, running commands, opening pull requests and working on tasks in the cloud.
Claude Code is Anthropic's agentic coding tool. It runs in your terminal, in VS Code and JetBrains, inside the Claude desktop app and in the browser, and it works the way a colleague with repository access would: reading and editing files across the codebase, running commands and tests, and opening pull requests. Sessions can move to the cloud and keep running with your laptop closed.
It is extended through skills, hooks, MCP servers and plugins, and it ships a public CHANGELOG with several releases a week. Access comes with Claude Pro, Max, Team and Enterprise plans, or through API usage.
Pricing model: Paid
Common questions about Claude Code
Claude Code is retiring fast mode for Opus 4.8. It was deprecated on October 30, 2026, and will be removed on November 30, 2026. After that, fast mode requests will return an error rather than fall back to standard Opus 4.8. The changelog recommends migrating to Opus 5.5.
Version 2.1.316 also adds CLAUDECODESUBAGENTEFFORT to set a default effort level for subagents, plus CLAUDECODEDISABLEULTRATHINK to turn off the ultrathink keyword.…
Claude Code 2.1.314 adds a setting to block its Write, Edit, and NotebookEdit tools from writing outside your working directories. Set permissions.blockWritesOutsideWorkingDirectories to apply the restriction in every permission mode.
In VS Code and Claude Desktop's Code tab, idle conversations of 200K or more tokens now get summarized. The next message uses that summary if the prompt cache has expired.
The release also fixes sessions repeatedly failing when an admin caps model effort.…
Claude Code 2.1.312 fixes a case where Claude could overwrite a very large CLAUDE.md, rules file or @-included file in a subdirectory after reading only its start. Write now asks Claude to read the whole file or edit it in place.
It also fixes missing managed settings and organization policy after VS Code sign-in, plus skills not being found after a Bash cd deeper into a git repository.
In fullscreen, you can now click the “to send now” hint under a queued message to send it immediately.…
Claude Code 2.1.311 fixes long freezes caused by very long list items in CLAUDE.md, plus stalled edits when changing thousands of lines.
Standard-context sessions that go over 200K tokens now compact and continue instead of stopping with “Usage credits required for 1M context.” Interrupts also work while an MCP resource read is waiting or a skill is downloading.
The release fixes subagents missing tools from MCP servers that connect after they start, and stale MEMORY.md and CLAUDE.md content…
Claude Code 2.1.310 adds the Monitor tool and plugin monitors on Bedrock, Vertex and Foundry, including when telemetry or nonessential traffic is disabled. Long foreground sleeps are now blocked in those setups too.
The release also fixes requests failing behind some gateways and background cleanup deleting recent files when cleanupPeriodDays is set very high. File read permission prompts now show where symbolic links lead.…
Claude Code 2.1.309 fixes several gaps in sandboxing and permission checks.
Sandboxed Bash commands on macOS and Linux no longer run shell startup files outside the sandbox. On macOS, denyWrite glob patterns now apply at a path's real location when the folder is reached through a symlink. Sandboxed commands and auto-accepted edits also can no longer modify the file supplied through --mcp-config.
The release also fixes permission checks for some compound Bash commands and closes gaps in…
Claude Code 2.1.307 adds /claude-api api-version to move Claude API code to API version 2026-10-21.
It also fixes HTTP MCP tool calls being sent twice when a connection closes mid-call. Only read-only or idempotent tools, which can safely be repeated, are now retried.
On macOS, a locked login keychain gets an explanation instead of a misleading “needs authentication” message. The next run after unlocking connects without a new sign-in.…
Claude Code 2.1.305 fixes a 2.1.304 regression that made every request fail on Vertex, Bedrock and other providers when tool search was rejected. It now retries without tool search.
The release also fixes new worktrees in bare repositories sometimes starting from local HEAD instead of the default branch, and makes cleanup of unchanged subagent worktrees reliable.
A new Linux setting, sandbox.allowNestedUserNamespaces, lets sandboxed commands create user namespaces for tools such as Chromium…
Claude Code 2.1.304 adds /auto-mode-setup, a guided setup that reads your project and recent sessions to teach auto mode about your environment. It also works on Bedrock, Vertex, Foundry, Claude apps gateway and telemetry-disabled sessions.
On Pro and Max, a warning now appears when resuming an inactive conversation would use a noticeable share of your 5-hour usage limit. You can resume anyway or start a new conversation and refer Claude to the previous one.…
Claude Code 2.1.303 brings bug fixes and reliability improvements, according to its changelog.
Original source
Other products to explore and follow.
The people behind Claude Code
Head of Product, Claude Code
Anthropic co-founder and President
Anthropic co-founder and Chief Compute Officer
Anthropic co-founder and Chief Science Officer
Anthropic co-founder and Chief Architect
Anthropic Chief Product Officer
A new Claude app where Claude works as a proactive agent: a morning brief, tasks you hand it, and work that continues when you step away.
Spotify's vendor-neutral agentic development environment — run and manage agent sessions across Claude Code, Gemini CLI, and Codex from one place.
Moonshot AI assistant for coding and agentic work. 3D Make connects to tools like Blender and CAD tools.
Cloud-based AI coding agent by OpenAI — assign tasks and Codex writes, tests, and ships code in a sandboxed environment.
Scan any site or MCP server and get a 0–100 agent-readiness score — can agents find you, access you, use you, and pay you.
AI code editor built on VS Code — agents that edit across your codebase, run in the cloud, review pull requests, and now host your repos.
