Claude Code for Enterprise
Steer what you ship
Hand your team the coding agent enterprises count on with access to real systems.
*Salesforce Ventures is an investor in Anthropic.
1 of 6: Muralidhar Krishnaprasad, and Richard Pack, Salesforce
“We intentionally let the model open PRs, review code, and query our logging systems and analytical databases, all to help people move faster. That’s where Anthropic’s built-in safeguards are really critical.”
~36%
of pull requests start as a prompt, with the work done in an isolated dev box
“Citi chose to leverage Claude as part of its AI powered Developer Platform because of its advanced planning and agentic coding capabilities, focus on safety and reliability, and compatibility with our workloads.”
79%
reduced time to market for new features (from 24 days to 5 days)
“Enterprise AI requires pairing frontier intelligence with inline runtime security and agent governance. Our collaboration with Anthropic helps to turn the promise of secure autonomous AI into an enterprise reality.”
“Our engineers are now able to execute fleet-wide migrations at a pace that simply wasn't possible before. It's a promising sign of how AI can reduce the complexity of maintaining large code bases and a clear signal of how AI is reshaping development velocity.”
89%
employee adoption, 800+ AI agents deployed, 10x app usage growth
“Claude Code is moving our team up a level: we decide what needs to happen, and smooth the process so it can build and verify end-to-end. A big part of my job now is to keep as many instances of Claude Code busy as possible.”
1.5x
increase in merged PRs among top 100 contributors, with 90%+ of engineers using Claude Code daily
Designed for your most ambitious work, with safety, privacy, and security built in.
Agentic, within limits you set
Claude Code helps you build features in natural language. It writes code, creates tests, and opens PRs so your team spends time on decisions instead of implementation. Your admins set its baseline permissions, and managed rules take precedence over the ones developers add.
Full context, clear boundaries
Claude reads your entire codebase, helping to make legacy migrations (opens in new tab) and cross-file debugging easier to manage. With the sandbox on, the operating system limits which files and sites Claude's commands can reach.
Reviews before risky actions
In auto mode, a separate model checks higher-risk actions before they run. In an Anthropic study of 1,053 paid testers in a test environment, people caught 13.6% of injected dangerous commands, and auto mode blocked 89% of the same commands. Auto mode reduces risk but doesn't eliminate it.
Frontier models with safeguards built in
Current Claude models are trained to decline destructive actions, tested before release, and documented in published system cards. On the Anthropic API, new models can reach your team automatically.
Deploy and manage Claude Code across your organization.
Contribution metrics
BetaSee usage patterns and productivity gains from your team in a single dashboard, including PRs and committed code written with Claude Code's help.
Server managed settings
Set tool permissions, file access, and MCP servers for your whole org from one admin panel, no MDM required. Managed rules take precedence over those developers add.
Monitoring and auditing
Export session, token, and cost metrics with OpenTelemetry. On Enterprise, the Compliance API lets you pull activity and audit logs into your SIEM.
Access and permissions
Control access with SSO. Enterprise adds role-based permissions, SCIM provisioning, tenant restrictions for personal accounts, and IP allowlisting.
Spend and model control
On Enterprise, set spend limits for your org, groups, or individuals, choose which models people can use, and cap effort levels by role.
Security Center
BetaCompare your security settings to our recommended configuration, where there is one, in Organization settings (Enterprise plans only).
Enterprise governance is built in with SOC 2 Type II, ISO 27001, and ISO 42001 certified. Claude Code for Enterprise is encrypted in transit and at rest.
Controls you set
Permissions, sandboxing (macOS, Linux and WSL2), and deny rules can be enforced through managed settings.
Models trained for safety
Current Claude models are trained to decline destructive actions and are tested before release.
Open communication
We post service incidents on our status page (opens in new tab), and share the methods behind our safeguards for transparency.
Hand off work from Slack
BetaTag @Claude in a thread with a bug report or a feature request. Claude reads the whole thread, works through the task in stages, and posts the pull request back in the thread. Everything stays in the channel history, where your team can see it.
Set recurring work to run on its own
BetaSave a prompt, your repositories, and your connectors once. Claude Code runs them in the cloud on a schedule, from an API call, or when a GitHub event fires. Each run is a full session you can open, follow, and steer.
Code Review on your pull requests
Code Review can check pull requests automatically, in the repositories you choose. The optional security guidance plugin reviews the code Claude writes as it goes, and Claude Security (beta) finds and suggests fixes for issues in the code you already have.
Team
Claude Code is included with Team plan. Includes self-serve seat management and additional usage at standard API rates.
From $20
Per seat / month, billed annually
- Claude Code with Code Review, scheduled tasks (beta), and contribution metrics (beta)
- Claude Cowork, Artifacts, and more
- Claude Tag (beta), Microsoft 365, and enterprise search
- SSO, central billing, and admin controls
- Mixed seat types and up to $500 a month in pooled API credits
Enterprise
Enterprise includes everything in the Team plan, plus advanced security, data, and user management.
All Team plan features, plus:
- Claude Security for managed vulnerability scanning (beta)
- Group spend limits and role-based permissions
- SCIM, audit logs, and the Compliance API
- Custom data retention controls
- Network-level access control and IP allowlisting
Usage limits (opens in new tab) apply. Prices shown don’t include applicable tax. Price and plans are subject to change at Anthropic's discretion.
Your admins set permissions, sandboxing, and deny rules through managed settings, and managed rules take precedence over the ones developers add. Claude is trained to resist hidden instructions and decline destructive actions. We post service incidents on our status page and share the methods behind our safeguards. Local sessions run on your machine and call the model API directly. Cloud sessions run in isolated virtual machines.
Claude Code deploys through Amazon Bedrock, Gemini Enterprise Agent Platform, or Microsoft Foundry, integrating with your existing IAM configurations, plus CloudTrail logging on AWS. You can export usage and activity to your observability stack with OpenTelemetry, and on Claude Enterprise the Compliance API lets security teams pull Claude Code session records from the CLI and desktop app for audits and eDiscovery.
Yes. Deploy CLAUDE.md files to system directories for company-wide standards, or create repository-specific files for project architecture and contribution guidelines. Claude Code references these automatically.
Everything in Team, plus advanced security, data, and user management, including: SCIM, role-based permissions, custom data retention, and IP allowlisting.
Support options vary by agreement. Contact sales for details.
Deploy through Anthropic Cloud (SaaS), Amazon Bedrock, Gemini Enterprise Agent Platform, and Microsoft Foundry. VPC isolation and private endpoints may be available through your cloud provider.