Anthropic pricing
Claude API cost and Anthropic API pricing
As of November 2026, the Claude API costs $2 per million input tokens and $10 per million output tokens on Claude Sonnet 5.5, from $0.10 and $0.50 on Haiku 5.5 up to $10 and $50 on Fable 5.1. Claude Enterprise is $20 a seat a month, billed annually, plus usage at API rates.
API pricing
How much does the Claude API cost per million tokens?
Anthropic prices the API per million tokens (MTok), with input and output priced separately and output five times input on every current model.
| Model | Input | Output | Cache write, 5 min | Cache hit | Batch input / output |
|---|---|---|---|---|---|
| Claude Fable 5.1 | $10 | $50 | $12.50 | $0.25 | $5 / $25 |
| Claude Opus 5.5 | $4 | $20 | $5 | $0.20 | $2 / $10 |
| Claude Sonnet 5.5 | $2 | $10 | $2.50 | $0.10 | $1 / $5 |
| Claude Haiku 5.5, prompts up to 100,000 tokens | $0.10 | $0.50 | $0.125 | $0.01 | $0.05 / $0.25 |
| Claude Haiku 5.5, prompts over 100,000 tokens | $0.50 | $2.50 | $0.625 | $0.05 | $0.25 / $1.25 |
US dollars per million tokens, standard global routing. Source: Anthropic, Pricing (Claude Platform docs), as of 3 November 2026.
The bill
What drives the Claude API bill?
The per-token rate is only the start. Five settings move the total, in both directions.
Prompt caching
A cache hit costs 0.1x the base input price, 0.05x on Opus 5.5 and Sonnet 5.5 and 0.025x on Fable 5.1. Writing the cache costs 1.25x for five minutes or 2x for one hour.
Batch processing
The Batch API takes 50% off input and output for requests that can wait for an asynchronous answer.
Long context
Claude 4.6 and later models, except Haiku 5.5, bill the full 1M-token context at the standard rate. Haiku 5.5 charges the whole request at the higher rate once the prompt passes 100,000 tokens.
Fast mode
Opus 5.5 in fast mode costs $8 input and $40 output per million tokens, twice the standard rate. It does not combine with the Batch API.
US-only inference
Setting inference to the US on Claude 4.6 and later models applies a 1.1x multiplier to every token category, cache reads and writes included.
Tools
Web search costs $10 per 1,000 searches plus the tokens it returns. Web fetch has no charge beyond the tokens it adds.
Where the surprises come from: Claude 4.7 and later models use a tokenizer that produces about 30% more tokens for the same text, so a migration can raise the token count before any rate changes. Older models cost more, not less: a key still pinned to Claude Opus 4.1 pays $15 and $75 per million tokens. Source: Anthropic, Pricing, as of 3 November 2026.
Seats
How much do Claude Team and Claude Enterprise cost?
Claude for work is sold by the seat. On Enterprise the seat is only part of the price: usage is billed on top at API rates.
| Plan | Price | Terms |
|---|---|---|
| Team, standard seat | $20 per seat a month billed annually, $25 billed monthly | For teams of 2 to 150 |
| Team, premium seat | $100 per seat a month billed annually, $125 billed monthly | 5x more usage than a standard seat |
| Enterprise | $20 per seat a month, billed annually, plus usage at API rates | 20-seat minimum; admins set user and organization spend limits |
| Pro (individual) | $17 a month with annual billing ($200 up front), $20 billed monthly | One person |
US dollars, excluding tax. Source: Claude, Plans and pricing, as of 3 November 2026.
Integration
Why we built a Claude API integration
Anthropic gives administrators good data about Anthropic. A budget owner needs the same data next to every other AI vendor and every other piece of software.
Why teams choose it
Anthropic describes Opus 5.5 as built for long-running agentic coding and knowledge work, and the 1M-token context at the standard rate suits long documents and large codebases.
What the vendor's tools show
The Claude Console tracks usage and cost, and Anthropic's usage and cost APIs cover both Console organizations and Claude Enterprise.
What they leave out
One vendor at a time. The OpenAI keys, the Copilot seats, the Cursor seats and the SaaS estate sit elsewhere, and an API key or workspace is not yet a person or a cost centre.
What our connector reads
- Anthropic Console: detailed token analysis and usage per user. Read by LCM and SIM.
- Anthropic Enterprise: Claude Enterprise seats and usage per user. Read by LCM and SIM.
In SIM
What does Claude spend look like in SIM?
Software Inventory Manager (SIM) puts Claude usage per user next to the rest of the software estate, and LICENSEWARE's LCM holds the contract side.
Cost per team
Token usage per user from the Anthropic Console connection. Add the users up by team or cost centre and follow the total month by month.
Seats against use
Claude Enterprise seats set against usage per user, so a paid seat nobody uses shows before the renewal.
Beside the estate
AI spend sits in the same inventory as SaaS subscriptions and installed software, so one review covers all three.
The contract side
LCM keeps the agreement, the seat count and the renewal date next to the usage, so a renewal starts from what was used.
More on AI cost management and every connection on data sources. Included in the Core plan, with no separate price.
Other AI cost pages: OpenAI API cost, Gemini API cost, Copilot license cost and Cursor pricing.
FAQ
Frequently asked questions
PriceHow much does the Claude API cost?
As of November 2026, Claude Sonnet 5.5 costs $2 per million input tokens and $10 per million output tokens. Haiku 5.5 starts at $0.10 and $0.50, Opus 5.5 is $4 and $20, and Fable 5.1 is $10 and $50.
Free tierIs the Claude API free?
No. Anthropic gives new users a small amount of free credits to test the API, and after that every token is billed at the published rates.
EnterpriseIs Claude Enterprise priced per seat or by usage?
Both. Claude Enterprise is $20 per seat a month, billed annually, with a 20-seat minimum, and usage is billed on top at API rates. Admins can set spend limits per user and for the organization.
SavingsHow can I lower Claude API costs?
Cache repeated prompts, since a cache hit costs a tenth of the input price or less. Send work that can wait through the Batch API at half price. Use the smallest model that does the job, and move keys off older models such as Opus 4.1, which cost more than Opus 5.5.
TokensWhy did my Claude token count go up after changing models?
Claude 4.7 and later models use a newer tokenizer that produces about 30% more tokens for the same text, according to Anthropic's pricing documentation. The rate per token can fall while the number of tokens rises.
Per userCan I see Claude API cost per user?
Yes. LICENSEWARE's Anthropic Console connector brings detailed token analysis and usage per user into SIM, and the Anthropic Enterprise connector adds Claude Enterprise seats and usage per user.
See your Claude spend per person
Connect the Anthropic Console or Claude Enterprise and SIM shows usage per user. Book 30 minutes and we walk through it with you.