The short answer: Claude Enterprise has a published seat price, but the seat is only part of the bill. Anthropic lists the Enterprise plan at US$20 per seat per month, billed annually, and states that the seat fee covers access only: every token your team uses in Claude, Claude Code and Cowork is billed separately at standard API rates, with no included allowance and no per-seat usage cap. Self-serve Enterprise needs at least 20 seats; sales-assisted Enterprise needs at least 50. If you want a fixed per-user price with usage included, the Team plan (2 to 150 users) is the alternative: $20 per Standard seat per month billed annually ($25 monthly) and $100 per Premium seat ($125 monthly).
This guide explains how Claude’s business plans are billed, what Enterprise adds over Team, the difference between self-serve and sales-assisted Enterprise, and how to estimate a budget when half the cost is usage. Prices are those Anthropic publishes on its pricing page and help center at the time of writing; check current pricing before you commit.
Claude plans for organizations at a glance
| Plan | Size | Published price | Usage | Key controls |
|---|---|---|---|---|
| Team, Standard seat | 2 to 150 users | $20 per seat per month billed annually; $25 monthly | Included, more than Pro, with limits | SSO, central billing, admin controls for connectors, enterprise search, no training on content by default |
| Team, Premium seat | Mixed with Standard seats | $100 per seat per month billed annually; $125 monthly | 5x Standard usage | Same as Standard |
| Enterprise | 20+ seats self-serve, 50+ sales-assisted | $20 per seat per month billed annually | All usage billed at API rates; no seat-level cap | Everything in Team plus SCIM, audit logs, Compliance API, custom retention, spend limits, IP allowlisting, HIPAA-ready option |
For comparison, Anthropic’s individual plans are Pro at $17 per month with annual billing ($200 upfront) or $20 monthly, and Max from $100 per month. Individual plans have no SSO, SCIM, audit logs or central billing, and Anthropic’s plan table shows model training as opt-out on Free, Pro and Max, while Team and Enterprise state no model training on your content by default. That difference is the main reason IT teams move staff off personal accounts.
How much is Claude Enterprise per month?
The fixed part is simple: US$20 per seat per month, billed annually, which is $240 per seat per year. Anthropic’s AWS Marketplace listing shows the same structure, with 20 Enterprise seats for one year at $4,800 plus usage fees based on Anthropic API rates.
The variable part depends on how much your people use Claude and which models they use. Anthropic’s help center puts it plainly: usage isn’t included in the seat fee, and every token used in chat, Claude Code or Cowork is billed at standard API rates on top of the seat cost. Larger models cost more per token than smaller ones, and long agentic coding sessions consume far more tokens than quick questions. So the honest answer to “how much per month” is: $20 per seat, plus whatever your usage adds, which you control with spend limits.
What does a Claude Enterprise seat include?
According to Anthropic’s pricing page and help center, each Enterprise seat includes:
- Claude on web, desktop and mobile, with projects, file creation, research and memory.
- Claude Code for software work and Cowork for multi-step tasks.
- Claude Design, Slides and Docs, Claude Science, and Claude for Microsoft 365 and Outlook.
- Enterprise search across connected company sources, and connectors for Google Drive, Gmail, Google Calendar, GitHub, Microsoft 365 and Slack.
- Enterprise deployment of the desktop app and admin controls for remote and local connectors.
- Claude Security, listed as a beta.
What it does not include is usage: the seat is the access fee, and the work people do is metered separately.
How Claude Enterprise billing works
- One seat type. Enterprise uses a single seat priced per user per month and billed annually. Each seat gives access to Claude on web, desktop and mobile, plus Claude Code and Cowork.
- Metered usage. There is no token allowance to divide among users. Each person’s usage is metered and billed to the organization at API rates, so one heavy user does not reduce what others can use.
- Spend limits. Admins can set limits at the organization level and for individual users. This is the main cost-control lever, and you should configure it before launch.
- How usage is paid. Self-serve plans buy credits upfront; sales-assisted plans are billed monthly in arrears.
Self-serve vs sales-assisted Enterprise
Anthropic says both routes have the same features and the same seat price. The differences are in purchasing:
| Term | Self-serve | Sales-assisted |
|---|---|---|
| How to buy | Online | Through Anthropic’s sales team |
| Minimum seats | 20 | 50 |
| Payment | Credit card, debit card or ACH | Credit card, ACH or invoicing (bank transfer only for invoices of $50,000 or more) |
| Currency | USD only | Multi-currency |
| Usage billing | Credits purchased upfront | Billed monthly in arrears |
| HIPAA-ready option and BAA | Available | Available |
Two practical notes from the help center: with ACH, the plan activates only once payment arrives (up to five business days), and auto-reload of credits needs a card on file. Claude Enterprise is also sold through AWS Marketplace, which some companies prefer because the spend counts against an existing AWS commitment.
What’s the difference between Claude Team and Enterprise?
| Capability | Team | Enterprise |
|---|---|---|
| Pricing model | Seat price with usage included (limits apply) | Seat price plus all usage at API rates |
| Size | 2 to 150 users | 20 or 50 seat minimum, no stated maximum |
| Single sign-on (SSO) | Yes | Yes |
| SCIM provisioning | No | Yes |
| Role-based access with fine-grained permissions | No | Yes |
| Audit logs and Compliance API | No | Yes |
| Custom data retention | No | Yes |
| Network-level access control and IP allowlisting | No | Yes |
| Customer-managed encryption keys | No | Yes |
| US-only inference | No | Yes |
| HIPAA-ready configuration with BAA | No | Available to eligible organizations |
| Spend limits by user and organization | No | Yes |
Team suits a department that wants predictable per-seat costs and basic admin. Enterprise suits organizations that need identity lifecycle automation, audit trails, retention control or regulated-data handling, and that are comfortable paying for usage as it happens.
What’s the difference between Claude Pro and Enterprise?
Pro is a personal subscription billed to one person, with usage limits and no organizational control. Enterprise is an organization account: the company owns the workspace, provisions users through SSO and SCIM, sees audit logs, sets retention, and pays for usage centrally. For employers, the data-use terms are the deciding factor: Anthropic’s enterprise page says prompts, data and results on its business plans are not used to train its models by default, while its pricing table lists training as opt-out on the individual plans.
Why is Claude Enterprise so expensive?
The seat fee is not high on its own; it is the same $20 per month as a Team Standard seat on annual billing. What makes some Enterprise bills large is the usage model: heavy users, especially developers running long Claude Code sessions or agents working through large codebases, can consume many tokens, and those tokens are billed at API rates with no cap unless you set one. Three things keep it under control:
- Set organization and per-user spend limits on day one.
- Review usage by user and team monthly through the analytics tools, and move light users to the right plan if you run more than one.
- Agree internally on which work justifies the largest models.
Are there other versions of Claude for enterprises?
Yes, in three senses. The Team plan is the smaller-business version. Organizations on older contracts may see legacy seat types: Anthropic’s help center says Enterprise organizations with Chat and Chat + Claude Code seats, or with Standard and Premium seats, keep those until their next renewal and then move to the single Enterprise seat model. And developers can build on Claude models directly through the Claude API, which is priced per token and is a different product from the Claude app plans.
How to estimate your Claude Enterprise budget
Annual cost = (seats x $20 x 12) + expected monthly usage x 12
The seat part is fixed. For a hypothetical 100-seat organization: 100 x $20 x 12 = $24,000 per year in seat fees. The usage part is where estimates go wrong, so build it from a pilot, not a guess:
- Run a four to eight week pilot with a mix of light users, daily writers and analysts, and developers.
- Record usage cost per user per week for each group.
- Multiply each group’s average by its headcount and by 52.
- Add a buffer for adoption growth and set spend limits at that level.
If the usage estimate makes the total unpredictable for finance, compare it with Team, where usage is included in the seat price up to the plan limits, or ask sales about the options available at your volume.
Enterprise requirements checklist
- Identity: SSO, SCIM provisioning, role-based permissions.
- Data use: no training on your inputs and outputs by default, per Anthropic’s enterprise page; link its Trust Center in your review.
- Certifications: Anthropic’s enterprise page lists SOC 2, ISO 27001, GDPR and CCPA compliance.
- Regulated data: HIPAA-ready configuration with a BAA for eligible Enterprise organizations.
- Audit and eDiscovery: audit logs and the Compliance API, which gives programmatic access to activity logs, chat histories and file content, filterable by user and time.
- Residency and encryption: US-only inference and customer-managed encryption keys are listed Enterprise features.
- Connectors: Google Drive, Gmail, Google Calendar, GitHub, Microsoft 365 and Slack, with admin control over which are allowed.
How Claude Enterprise compares on price
| Product | Published entry price | Usage |
|---|---|---|
| Claude Enterprise | $20 per seat per month, billed annually | Billed separately at API rates |
| ChatGPT Enterprise | Quote only | Credits or token billing beyond included limits |
| Microsoft 365 Copilot | $30 per user per month, paid yearly | Included for in-app use; agent credits metered |
| Gemini Enterprise Standard/Plus | From $30 per seat per month | Included within quota; pay-as-you-go edition rolling out |
More detail: ChatGPT Enterprise pricing, Microsoft Copilot pricing and Gemini Enterprise pricing. You can also read our Claude product profile.
How to choose
- Fewer than 20 users: Team is the only business option; Enterprise self-serve starts at 20 seats.
- You need SCIM, audit logs, retention control or a BAA: Enterprise.
- Finance needs a fixed number: Team, or Enterprise with strict spend limits and a pilot-based usage budget.
- Developers are your heaviest users: Enterprise includes Claude Code in every seat, but model usage cost is where your budget will move; pilot it first.
- You need invoicing or non-USD billing: sales-assisted Enterprise (50+ seats).
If developers are your heaviest users, read our enterprise guide to AI coding assistants. For the wider picture, see enterprise AI platforms and AI governance tools.
Frequently asked questions
What is Claude Enterprise?
Anthropic’s plan for organizations that need advanced security and compliance. It includes everything in Team plus SCIM, audit logs, the Compliance API, custom retention, spend limits and a HIPAA-ready option, with a per-seat fee and usage billed at API rates.
How much is Claude Enterprise per user?
US$20 per seat per month, billed annually, plus the usage each user generates at API rates.
Is there a minimum number of seats?
Yes: 20 seats for self-serve Enterprise and 50 for sales-assisted. Team works for 2 to 150 users.
Does Claude Enterprise have usage limits?
Usage-based Enterprise plans have no plan or seat-level limits. You pay for what is used, and admins cap spend with organization and user limits. Older seat-based Enterprise contracts with Standard and Premium seats still have limits.
Does Anthropic train on Claude Enterprise data?
Not by default. Anthropic’s enterprise page says inputs and outputs are not used to train its models by default and points to its Trust Center for details.
Can Claude Enterprise be used with health data?
Eligible organizations can enable a HIPAA-ready configuration and accept a BAA from organization settings, on both self-serve and sales-assisted plans.
Is Claude Code included in Enterprise?
- Yes, every Enterprise seat includes Claude Code and Cowork.
- The tokens they use are billed at API rates like any other usage.
Related Articles
AI Software
Fireflies vs Otter 2026: Pricing, Limits and Which to Pick
Continue reading →
AI Software
AI in Finance: 12 Use Cases for Corporate Finance Teams in 2026
Continue reading →
AI Software
Copilot vs ChatGPT for Business (2026): Price, Security and Use Cases
Continue reading →
AI Software
AI in Procurement in 2026: 10 Use Cases, Tools and How to Start
Continue reading →
