Claude Enterprise from Anthropic lists at $60 per seat per month on annual commit and $25 per seat per month on the Team tier, with realised pricing of $42 to $54 per seat per month at deal sizes above 1,000 seats on three-year commit. Claude Pro for individuals sits at $20 per month. The API price list runs $15 per million input tokens and $75 per million output tokens for Claude Opus 4 (the frontier model), $3 input and $15 output for Claude Sonnet 4, and $0.25 input and $1.25 output for Claude Haiku 4. The hidden cost layer for Anthropic procurement is the Computer Use Tool consumption (billed per agent action) and the Projects feature consumption (billed against the team's shared context pool). A 1,000-seat Claude Enterprise rollout with moderate API usage lands at $620,000 to $780,000 per year all-in.
Inside This Pillar
- Claude Enterprise 2026 list pricing
- Enterprise vs Team vs Pro vs Free
- Anthropic API per-token pricing
- Projects feature pricing model
- Computer Use Tool pricing
- Claude on Bedrock vs Vertex vs direct Anthropic
- Enterprise security, SSO, audit, retention
- IP indemnity and the Anthropic terms
- Realised discount bands by seat count
- 1,000-seat and 10,000-seat rollout cost model
- The five negotiation levers that work in 2026
Claude Enterprise 2026 list pricing
Anthropic publishes Claude seat pricing across four named tiers in 2026: Free, Pro ($20 per user per month), Team ($25 per user per month with annual billing and a $30 monthly variant), and Enterprise (custom, with $60 per seat per month as the typical published reference). Enterprise is the only tier that includes SAML SSO, SCIM provisioning, custom retention windows, the zero-retention commitment for prompts and completions, and the contractual indemnity language most enterprise buyers require.
| Tier | List per user per month | Context window | Models available |
|---|---|---|---|
| Claude Free | $0 | 200K tokens | Claude Sonnet 4 |
| Claude Pro | $20 | 200K tokens | Sonnet 4 + Opus 4 (limited) |
| Claude Team (annual) | $25 | 200K tokens | Sonnet 4 + Opus 4 |
| Claude Team (monthly) | $30 | 200K tokens | Sonnet 4 + Opus 4 |
| Claude Enterprise | $60 (typical, custom) | 500K tokens (expanded) | All models including Haiku 4 |
A 1,000-seat Claude Enterprise rollout at $60 per seat lists at $720,000 per year. Realised pricing on a three-year commit at 1,000 seats typically lands between $42 and $54 per seat per month, or $504,000 to $648,000 per year. Discount realisation depends on TCV, term length, and the credible threat of ChatGPT Enterprise or Microsoft 365 Copilot in the deal.
Why the expanded context window matters: Claude Enterprise extends the context window from 200,000 tokens (the standard window across Pro and Team) to 500,000 tokens. For most knowledge-work use cases the 200K window is sufficient. For document-heavy workflows (legal review, contract analysis, large codebase navigation, multi-document research), the 500K window is a step-change in capability that frequently justifies the Enterprise tier on its own. The 500K window holds approximately 1,200 pages of dense text, which covers the typical 10-K filing plus the prior year's 10-K and analyst transcripts in a single context.
Enterprise versus Team versus Pro versus Free
The tier comparison determines what the customer gets for the unit price. The Enterprise tier delivers the controls that enterprise security, legal, and compliance teams require for production deployment.
| Feature | Pro | Team | Enterprise |
|---|---|---|---|
| SAML SSO | No | No | Yes |
| SCIM provisioning | No | No | Yes |
| Custom data retention | No | Limited | Yes (configurable) |
| Zero-retention default | No | No | Yes |
| Admin console + audit log | No | Basic | Full |
| Context window | 200K | 200K | 500K |
| Projects (shared context) | Personal only | Team-level | Org-level with access controls |
| Computer Use Tool | No | No | Yes (consumption-billed) |
| Custom integrations (MCP) | Limited | Yes | Yes (priority support) |
| HIPAA BAA | No | No | Yes |
| Priority support SLA | No | No | Yes |
The most common enterprise rollout pattern is to deploy Team to low-risk broad-population users (sales, marketing, general office) and Enterprise to high-risk knowledge work (legal, finance, research, engineering). The tier delta on a per-seat basis is $35 per month, which is materially less than the comparable ChatGPT Business-to-Enterprise gap on most negotiated outcomes.
Anthropic API per-token pricing
Most Claude Enterprise customers also consume the Anthropic API for embedded AI features in custom applications. API pricing is per-token and runs separately from the Claude seat pricing. The 2026 token prices for the production Claude 4 model family are listed below.
| Model | Input per 1M tokens | Output per 1M tokens | Context window |
|---|---|---|---|
| Claude Opus 4 | $15.00 | $75.00 | 500K |
| Claude Sonnet 4 | $3.00 | $15.00 | 500K |
| Claude Haiku 4 | $0.25 | $1.25 | 200K |
| Claude Opus 4 (Batch) | $7.50 | $37.50 | 500K |
| Claude Sonnet 4 (Batch) | $1.50 | $7.50 | 500K |
| Cached input (1-hour) | 10% of standard | n/a | n/a |
| Cached input (5-minute write) | 125% of standard | n/a | n/a |
Prompt caching is the largest single optimisation on the Anthropic API. Workloads with stable system prompts (typical for RAG, chat applications, agent frameworks) typically achieve 60 to 85 percent cost reduction on cached input tokens. Batch API pricing at 50 percent of standard applies to non-real-time workloads (overnight document processing, periodic embedding refresh, evaluation runs).
Projects feature pricing model
Projects is the Claude Enterprise feature that lets a team share context, instructions, and knowledge files across users. Projects are included in Team and Enterprise tiers with consumption limits on the shared context pool. The Enterprise tier provides organisation-level Projects with role-based access controls and audit logging, which Team-tier does not.
The Projects context pool is bounded by the tenant's allocated capacity. Enterprise customers receive a meaningful default pool that covers typical knowledge-work usage. Heavy use cases (full corporate knowledge bases, large code repositories, multi-thousand-document research collections) may consume the pool faster than expected, triggering overage billing or expansion conversations. The pricing structure for overage is negotiated per-tenant rather than published.
Computer Use Tool pricing
The Computer Use Tool lets Claude operate a virtual computer interface (screenshots, mouse, keyboard) to execute multi-step tasks in browser and desktop applications. Computer Use is billed per token (same Opus or Sonnet token rates) plus a per-action overhead for the screenshot and execution loop. Typical agent run costs in 2026 land between $0.15 and $0.85 per completed task depending on task complexity and the model variant used.
Computer Use is the highest-velocity new spend category in 2026 Anthropic enterprise contracts. The use cases that justify the cost are narrow but high-value: data entry into legacy systems without APIs, multi-step research and synthesis, vendor portal interaction, and similar repetitive browser-bound workflows. The risk is unbounded agent activity that runs up large token bills against vague exploratory prompts. The mitigation is per-team budget caps and run-time limits enforced through the admin console.
Claude on Bedrock versus Vertex versus direct Anthropic
Claude models are available through three distribution channels: direct Anthropic API, AWS Bedrock, and Google Vertex AI. The token economics are identical across channels (Anthropic does not pass discount through cloud sub-channels). The procurement decision turns on contract structure, deployment integration, and indemnity terms.
| Channel | Contract vehicle | Data residency | Best for |
|---|---|---|---|
| Anthropic direct | Anthropic enterprise MSA | US, EU expansion in 2026 | Anthropic-first organisations, full feature access |
| AWS Bedrock | Burns AWS EDP commit | Per AWS region | AWS-committed estates, regional flexibility |
| Google Vertex AI | Burns Google Cloud commit | Per GCP region | GCP-committed estates, Gemini comparison workloads |
For AWS-committed customers with material EDP burn, Claude on Bedrock is the right answer because the token spend reduces the EDP commitment to be earned out. See AWS Bedrock pricing 2026 and AWS EDP pillar for the EDP burn-down economics. For GCP-committed customers the equivalent applies through the Vertex AI route. For organisations without material hyperscaler commits, direct Anthropic is typically simpler.
The MCP advantage: Claude was the first frontier model to natively support the Model Context Protocol (MCP), which standardises how the model connects to external tools, data sources, and applications. Enterprise buyers integrating Claude with existing systems (Slack, Jira, Salesforce, internal databases) gain meaningful integration velocity from the MCP ecosystem versus equivalent custom integration work on other models. MCP is open standard, so this is not a permanent moat, but it is a real 2026 advantage when scoping enterprise rollouts.
Enterprise security, SSO, audit, retention
The Enterprise tier includes the controls that enterprise security and legal teams require for production deployment. SAML 2.0 SSO with major identity providers (Okta, Entra ID, Ping, OneLogin), SCIM 2.0 automated user provisioning and deprovisioning, configurable data retention from one day to indefinite, default zero-retention for prompts and completions (the model does not train on customer data), full admin console with usage analytics, audit log with API export, HIPAA BAA availability, SOC 2 Type II report, ISO 27001 certification, and the Anthropic Acceptable Use Policy with named carve-outs.
IP indemnity and the Anthropic terms
Anthropic publishes a copyright indemnity for Claude output as part of the standard Enterprise MSA. The indemnity covers direct copyright infringement claims arising from Claude-generated output, with the standard exclusions for misuse, fine-tuning on infringing data, and prompts designed to elicit infringing content. The indemnity is structured similarly to the Microsoft and OpenAI commercial-output indemnities.
The procurement question is how the indemnity caps and notice provisions compare across vendors. Anthropic, OpenAI, and Microsoft have meaningfully different cap structures and customer-side obligations that affect the practical value of the indemnity. The comparison across the three vendors is documented in our AI contract data residency and IP rights guide.
Realised discount bands by seat count
Claude Enterprise discount realisation in 2026 follows a similar shape to ChatGPT Enterprise, with Anthropic typically slightly more flexible at the higher-seat bands as the company expands its enterprise sales motion.
| Seat count band | Annual commit price | Three-year commit price |
|---|---|---|
| 250 to 500 seats | $55 to $62 per seat per month | $48 to $56 |
| 500 to 1,500 seats | $48 to $58 | $42 to $52 |
| 1,500 to 5,000 seats | $42 to $52 | $36 to $46 |
| 5,000 to 15,000 seats | $36 to $46 | $30 to $40 |
| 15,000+ seats | $30 to $40 | $24 to $34 |
1,000-seat and 10,000-seat rollout cost model
| Cost line | 1,000-seat rollout | 10,000-seat rollout |
|---|---|---|
| Enterprise seat licences (3yr commit) | $540,000 | $3,600,000 |
| API consumption (embedded apps) | $60,000 to $200,000 | $500,000 to $1,800,000 |
| Computer Use Tool (10% attach, moderate use) | $15,000 to $40,000 | $150,000 to $400,000 |
| Projects context pool overage (typical) | $5,000 to $20,000 | $50,000 to $200,000 |
| Total annual | $620K to $800K | $4.3M to $6.0M |
The five negotiation levers that work in 2026
- ChatGPT Enterprise quote in hand. Anthropic sales flexibility correlates directly with documented OpenAI alternative pricing at equivalent seat counts. See ChatGPT Enterprise pricing 2026.
- Hyperscaler channel arbitrage. If you have meaningful AWS EDP or GCP commit, validate Bedrock or Vertex routing to consume from the existing commit. The token price is identical but the commercial accounting changes materially.
- Three-year commit for unit price relief. Anthropic, like OpenAI, accepts meaningful unit-price reduction in exchange for three-year terms.
- Phased deployment with quarterly add windows. Commit at 65 to 75 percent of forecast initial seats with quarterly expansion at the contracted unit price.
- Bundle the API consumption alongside seats. Anthropic account teams have authority to discount API consumption in exchange for seat commit, especially for embedded application use cases.
The full Anthropic procurement framework lives in our enterprise AI vendor selection framework. For cross-vendor token comparison see enterprise LLM cost comparison. For the per-token negotiation pattern see AI usage-based pricing negotiation. For broader counsel see AI procurement advisory, AI procurement guide, software licensing advisory, and cloud contract negotiation.