Pricing at a glance
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Decision point | What Anthropic states |
|---|---|
API billing | Per token by model, billed on actual monthly usage in USD. Credit cards for standard accounts; invoicing for Enterprise customers. |
Trying the API | New users get a small amount of free credits. |
Annual billing | Offered on Pro and Team, billed once up front; Enterprise is annual only. |
Tax | Prices exclude applicable tax and may change at Anthropic's discretion. |
Which plan fits
Daily individual work or coding: Pro; Claude Code shares the plan's usage pool with chat.
All-day sessions: Max or Ultra.
2 to 150 people needing SSO: Team; Premium seats for heavy users.
SCIM, audit logs or HIPAA-ready needs: Enterprise.
Building an application: the API.
Plans and API rates
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Plan | Published price | What it includes |
|---|---|---|
Free | $0 | Limited usage; chat on web, desktop and mobile; web search, files and code; memory; connectors; Artifacts; Claude Design, Slides and Docs; up to 5 Projects; Sonnet and Haiku, not Opus. |
Pro | $17 per month with annual subscription discount ($200 billed up front); $20 if billed monthly | Everything in Free, plus more usage, Opus, Claude Code, Claude Science, Research, Projects, scheduled tasks, Chrome and Microsoft 365; Fable paid with usage credits. |
Max | $100 per month, billed monthly | Everything in Pro, plus 5x Pro usage every 5 hours and every week; Fable up to half the weekly limit; $100 a month in API credits; priority access at high traffic times. |
Ultra | $200 per month, billed monthly | Everything in Max, plus 10x Pro usage every week; no 5-hour limit for verified accounts; $200 a month in API credits. |
Team Standard seat | $20 per seat / month if billed annually; $25 if billed monthly | More usage than Pro; Claude Code and Cowork; SSO; central billing; project sharing; no training on your content by default; up to $500 per month in pooled API credits. |
Team Premium seat | $100 per seat / month if billed annually; $125 if billed monthly | 5x more usage than Standard seats; Fable for 50% of weekly limits. |
Enterprise | US$20/seat/month, billed annually, 20 seat minimum, plus usage at API rates | All Team features, plus spend limits, role-based access, SCIM, audit logs, Compliance API, custom data retention, IP allowlisting and a HIPAA-ready offering. Self-serve or via sales. |
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Model | Input / MTok | Output / MTok | Cache write / MTok | Cache read / MTok |
|---|---|---|---|---|
Claude Opus 5.5 | $4 | $20 | 1.25x input | $0.20 |
Fable 5 (legacy) | $10 | $50 | $12.50 | $1 |
Opus 5, 4.8, 4.7, 4.6, 4.5 (legacy) | $5 | $25 | $6.25 | $0.50 |
Sonnet 4.6, 4.5 (legacy) | $3 | $15 | $3.75 | $0.30 |
Sonnet 5 (legacy) | $2 | $10 | $2.50 | $0.20 |
Haiku 4.5 (legacy) | $1 | $5 | $1.25 | $0.10 |
Cache hits cost $0.10 per MTok on Claude Sonnet 5.5. Fast mode costs $8 input and $40 output per MTok on Claude Opus 5.5, and $10 and $50 on Claude Opus 5 / Claude Opus 4.8 (research preview). Fast mode is not available with the Batch API.
How the bill adds up
Individual and Team plans are flat fees per user or seat, bounded by usage limits: for most accounts a rolling five-hour window, plus weekly limits on paid plans. At a limit, paid users can turn on usage credits at standard API rates. Enterprise adds API-rate usage to its seat fee. The API bills per-token model rates plus server-side tools such as web search.
Anthropic's worked example
One-hour Managed Agents session on Claude Opus 5.5: 50,000 × $4 / 1,000,000 input = $0.20, 15,000 × $20 / 1,000,000 output = $0.30, and 1.0 hour × $0.08 runtime = $0.08: $0.58 USD. With 40,000 input tokens as cache reads: 10,000 × $4 / 1,000,000 = $0.04, 40,000 × $4 × 0.05 / 1,000,000 = $0.008, plus the same output and runtime: $0.428 USD. Both totals exclude web search ($10 per 1,000 searches), the 1.1x US-only multiplier and fast mode pricing.
What moves the API bill
Caching. A 1-hour cache write costs 2x base input; reads cost 0.1x (0.05x on Opus 5.5 and Sonnet 5.5; 0.025x on Fable 5.1 and Mythos 5.1).
Batch. Anthropic states it saves 50%, and it combines with caching.
US-only inference. Adds a 1.1x multiplier on Claude 4.6 and later.
Costs that are easy to miss
Tokenizer. Claude 4.7 and later produce approximately 30 percent more tokens for the same text than earlier models.
Tool prompts. Tools add system prompt tokens, such as 286 on Opus 5.5 with tool choice auto or none.
Code execution. The pricing page lists 50 free hours daily per organization; the API docs list 1,550 free hours per month. Beyond that, $0.05 per hour per container. The docs add a 5-minute minimum and billing when files are attached.
Managed Agents. Runtime adds $0.08 per session-hour.
Refunds. Generally non-refundable, except under the Consumer Terms, local law or the EEA and UK 14-day withdrawal period.
Our take
If your team works in the Claude apps, choose by usage limits and admin needs; higher tiers mostly buy headroom. If you build on the API, model choice, caching and batch use shape the bill.
Evaluate this price
Comparing alternatives? Model this workload across OpenAI, Anthropic and Gemini API. Preview cost comparison →