Claude Team's published prices, checked on 18 October 2026, are $20 per Standard seat per month billed annually or $25 billed monthly, and $100 or $125 for Premium seats. Anthropic does not publish a token count for the plan. Light review, drafting and triage fit Standard seats, while heavy agent work needs Premium seats or usage credits.
What does the plan cost and include?
This section reflects Anthropic's pricing page and its Team plan help article, both fetched on 18 October 2026. Prices and plans are subject to change at Anthropic's discretion, so check the pages before you decide. Prices shown are for US customers and exclude tax.
- Standard seat: $20 per member per month billed annually, or $25 billed monthly.
- Premium seat: $100 per member per month billed annually, or $125 billed monthly.
- Seats: the Team plan needs at least two members and supports up to 150 seats. Larger organizations move to Enterprise.
- Included: the help article says Standard seats get more usage than Pro, access to all available models, Claude Code, Cowork, projects, connectors to tools such as Google Drive, GitHub, Microsoft 365 and Slack, admin and billing tools, single sign-on with domain capture, just-in-time provisioning, role-based permissions and spend controls.
Audit logs, SCIM and the compliance API appear in the Enterprise column on the pricing page and not in Team. If you need them for a security program, read the current comparison table.
What are the usage limits?
The help article says limits are per member, not shared across the team. If one person reaches their limit, others are unaffected, and limits cannot be transferred. Weekly limits reset at a fixed time assigned to each account. Standard seats include 1.25 times the Pro plan's per-session usage allowance, with a weekly limit across all models. Premium seats include 6.25 times the Pro per-session allowance. An organization can enable usage credits so members can keep working with Claude, Cowork and Claude Code after reaching included limits.
What the pages do not say is how many tokens that is. The pricing page expresses usage relative to other plans, with a footnote that usage limits apply. So any claim of "X tokens per month" for a Team seat should be treated as unverified. Measure your own usage in Settings, where the help article says your usage and next reset time appear.
Which security tasks suit a Standard seat?
Tasks with modest, bursty usage:
- Summarizing a vulnerability advisory and drafting a ticket.
- Reviewing a single pull request diff for obvious issues.
- Drafting a security policy or an incident runbook from your notes.
- Turning an audit request list into a checklist.
- Explaining a scanner finding to a developer.
Which tasks push toward Premium or credits?
Long agentic sessions use far more than a chat. Running a coding agent across a big repository, doing a multi-hour code review, or processing many documents in one sitting can consume a seat's allowance quickly. Give those users Premium seats, or enable usage credits for them. Mixing seat types is allowed, so you do not need to upgrade everyone.
How does Safeguard fit?
Safeguard publishes a guide for connecting its MCP server to Claude, with a custom connector called Safeguard Security. Once connected, Claude can query vulnerabilities, SBOMs, packages, licenses and findings that your Safeguard role allows, and the Safeguard side of the work has no per-call charge, since MCP tool calls are unlimited on every Safeguard plan. Claude's own usage still counts against your Anthropic plan. The two are separate: you connect Safeguard to Claude through MCP, and Anthropic's plan terms govern Claude.
Try the walkthrough at /demos/claude-with-safeguard.
What should you keep out of any assistant?
Check your data rules before pasting anything. Do not paste live credentials, customer personal data or unreleased vulnerability details into a tool unless your policy allows it. The pricing page says Team plans do not train on your content by default, but your own policy and contracts still decide what is acceptable.
What to do
- Fetch the current pricing and help pages and check the numbers against this post.
- List the security tasks you want help with and sort them into light and heavy.
- Start with Standard seats for light work, and watch usage for a month.
- Give Premium seats or credits only to people who hit limits.
- Set spend controls and roles in the admin settings.
- Write a short rule on what data may go into the assistant.
- Connect Safeguard through MCP using OAuth, and keep tool access to the minimum your team needs.
- Review cost and value every quarter.
Quick check
How many tokens does a Team seat include?
Anthropic's published pages do not give a token number. They describe usage relative to the Pro plan and say limits apply. Check your own usage in Settings.
Are usage limits shared across the team?
No. The help article says limits are per member, and cannot be transferred or shared.
Sources
- Claude pricing, fetched 18 October 2026
- Claude Help Center: What is the Team plan? (fetched 18 October 2026)
- Safeguard docs: MCP Integrations
Last reviewed: 18 October 2026.
