RSS
Tips

Claude Code pricing (November 2026): plans and API costs

What Claude Code costs in November 2026: which Claude plans include it, how much usage each one gives, API token prices for Fable 5.2, Opus 5.5, Sonnet 5.5 and Haiku 5.5, worked cost examples, and how to pick a plan.

Black calculator beside black pen on white printer paper illustrating Claude Code pricing (November 2026)
Photo by Mediamodifier on Unsplash

Updated November 20, 2026: Rechecked against Anthropic's pages on November 20: added 1-hour cache prices, corrected cloud-provider billing, and clarified usage-credit and API-credit rules.

Last checked: 2026-11-20. Every price here comes from Anthropic's own pages: the claude.com pricing page, the Claude API pricing docs, the Claude Code docs and the Claude Help Center. The plan and token-price tables link the page each figure comes from, and the full source list is at the end. Prices shown on claude.com exclude applicable tax (the Help Center notes that some regions include tax in the displayed price), and Anthropic says plans and prices can change.

Claude Code is included in every paid Claude plan, so there is no separate price for it. You pay in one of two ways. With a paid subscription (Pro, Max, Ultra, Team or Enterprise), your terminal work draws from the same usage pool as your chats. With a Claude Console account you pay per token at Anthropic's API rates. Through Amazon Bedrock or Google Cloud (Google managed), the cloud provider sets its own prices; Microsoft Foundry bills through the Azure Marketplace in Claude Consumption Units. The free Claude plan does not include Claude Code.

How much does Claude Code cost on each plan?

The cheapest plan with Claude Code is Pro at $20 a month ($17 a month billed annually). Max costs $100 and Ultra $200 a month, Team seats cost $20 or $100 per seat a month billed annually, and Enterprise costs $20 per seat a month plus usage at API rates.

Claude plans and Claude Code, as listed on November 20, 2026, before tax. US web prices; the Help Center says mobile prices may vary.
PlanPriceClaude Code usageSource
Free$0Claude Code not includedclaude.com/pricing
Pro$20 a month, or $17 a month billed annually ($200 up front)Pro usage every 5 hours and every week; Fable only through usage creditsclaude.com/pricing
Max$100 a month, monthly only5x Pro every 5 hours and every week; Fable up to half your weekly limitWhat is the Max plan?
Ultra$200 a month, monthly only10x Pro every week; no 5-hour limit for verified accounts (a temporary limit applies until then); Fable up to half your weekly limitWhat is the Ultra plan?
Team, Standard seat$20 per seat a month billed annually, $25 monthlyMore usage than Pro (no multiplier published)claude.com/pricing
Team, Premium seat$100 per seat a month billed annually, $125 monthly5x a Standard seat; Fable up to half the weekly limitclaude.com/pricing
Enterprise$20 per seat a month billed annually, 20-seat minimum, plus usage at API ratesUsage-based plans have no per-seat usage limits; admins set spend limitsTeam and Enterprise guide
Console API keyPay per tokenNo subscription usage cap; tier-based API rate limits applyModels, usage and limits

Team plans cover 2 to 150 people and include up to $500 a month in pooled API credits. Ultra is the former Max 20x plan under a new name: the Help Center says existing Max 20x subscribers moved to Ultra automatically with no price change. Annual billing is offered for Pro and Team; Max and Ultra are monthly only, and Enterprise is annual only. Ultra accounts are verified automatically from signals such as account and payment history; if Settings > Usage shows no Current session bar, the account is verified.

What do Claude models cost per token?

On the API, Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, Opus 5.5 costs $4 and $20, Sonnet 5.5 $2 and $10, and Haiku 5.5 $0.10 and $0.50 for prompts up to 100,000 tokens. When Claude Code signs in with an API key, each request is billed at these rates. The Claude Code model docs say the default model is Opus 5.5 on Pro, Max, Ultra, Team, Enterprise and the Anthropic API. Fable models are never the account default on any plan; you choose one with /model fable.

US dollars per million tokens (MTok) at standard speed on the Claude API, from Anthropic's API pricing page on November 20, 2026. Our API pricing calculator uses the same figures.
ModelInput $/MTokOutput $/MTok5-min cache write $/MTok1-hour cache write $/MTokCache hit $/MTok
Fable 5.2$8$40$10$16$0.20
Opus 5.5$4$20$5$8$0.20
Sonnet 5.5$2$10$2.50$4$0.10
Haiku 5.5, prompts up to 100K tokens$0.10$0.50$0.125$0.20$0.01
Haiku 5.5, prompts over 100K tokens$0.50$2.50$0.625$1$0.05

Several modifiers on the same pricing page change the final bill:

  • Fast mode costs twice the standard rate, $8 input and $40 output per MTok for Opus 5.5 and $16 and $80 for Fable 5.2, in exchange for up to 2.5x faster output.
  • A 1-hour cache write costs 2x the base input price, against 1.25x for the 5-minute cache.
  • US-only inference adds a 1.1x multiplier on Claude 4.6 and later models on the Claude API.
  • The 1M-token context window is billed at standard rates on Claude 4.6 and later models, except Haiku 5.5, which charges the higher rates above once a prompt passes 100,000 tokens.
  • Claude 4.7 and later models use a newer tokenizer that Anthropic says produces about 30% more tokens for the same text, so per-token prices are not directly comparable with older models.
  • The Batch API gives a 50% discount on input and output tokens for asynchronous jobs. It does not apply to Managed Agents sessions, and fast mode is not available with it.

For your own token mix, our Claude API pricing calculator turns these rates into a cost per request, per day and per month.

How much does a typical Claude Code session cost?

Anthropic's Claude Code cost docs say that across enterprise deployments Claude Code averages about $13 per developer per active day and $150 to $250 per developer per month, and 90% of users stay below $30 per active day. Your own figure depends on the model, the size of the codebase and how many sessions and subagents you run.

To show how the price table turns into a bill, here is one long session worked through by hand. The token mix is our own illustration, not a measurement. The cost docs say Claude Code sends your full conversation with every request and re-reads that history at the cached rate, so we assume most input comes from the prompt cache: 3 million cache-hit tokens, 300,000 tokens written to the 5-minute cache and 60,000 output tokens.

Our arithmetic from Anthropic's list prices for an assumed token mix. Illustrative only; real sessions vary.
ModelCache hits, 3MCache writes, 300KOutput, 60KTotal
Sonnet 5.5$0.30$0.75$0.60$1.65
Opus 5.5$0.60$1.50$1.20$3.30
Fable 5.2$0.60$3.00$2.40$6.00

Two things follow from the table. Output and cache writes, not cache hits, drive most of the cost, and moving from Sonnet 5.5 to Fable 5.2 more than triples the bill for the same work. In Claude Code, run /usage to see the current session's tokens and an estimated dollar figure computed at list price. The docs say subscribers can ignore that figure for billing, because their usage is included in the plan; they see plan usage bars on the same screen.

Is a subscription or the API cheaper for Claude Code?

For steady daily use, a subscription is usually cheaper, because the flat fee covers your usage up to the plan's limits. As a rule of thumb (our own, not Anthropic's): if you would spend more than $20 a month at API rates and Pro's limits cover your work, Pro costs less. If your use looks like Anthropic's enterprise average of $150 to $250 a month, Max at $100 or Ultra at $200 fixes the price of the usage the plan includes. With usage credits turned on, work past a limit is billed on top of the fee at API rates, up to the spend limit you set.

The API makes more sense when you need exact per-token accounting, run Claude Code in CI or scripts, or do not want a subscription usage pool; API keys still have tier-based rate limits. Organizations that use Amazon Bedrock or Google Cloud pay those providers' prices instead; Anthropic's pricing page links their price lists.

One trap catches subscribers: if an ANTHROPIC_API_KEY environment variable is set and you approve it, Claude Code uses that key instead of your subscription, and the Help Center warns this results in API usage charges. The authentication docs say to run /status to confirm which login is active.

Which Claude Code plan should you pick?

Pick by how much you code and how you want to be billed:

  • Pro, $20: you are trying Claude Code or work in a few focused sessions a week. Fable is available only as pay-as-you-go through usage credits.
  • Max, $100: you code most days and keep hitting Pro's limits. You get 5x Pro's usage, Fable within half of your weekly limit, and priority access at high-traffic times.
  • Ultra, $200: you run long or autonomous tasks. A verified account has no 5-hour limit, only a weekly limit of 10x Pro and the separate Fable limit.
  • Team: 2 to 150 people who need central billing, SSO and admin controls. Give the heaviest coders Premium seats.
  • Enterprise: you need SCIM, audit logs, role-based access or custom data retention, and are happy to pay for usage at API rates on top of the seat.
  • Console API key: CI jobs, automation, or a team that wants strict per-token billing with no subscription pool.

What are usage credits and monthly API credits?

Usage credits let Pro, Max and Ultra subscribers keep working after they hit a plan limit, billed separately at standard API rates. Requests continue until you reach the monthly spend limit you set, or until your balance runs out with auto-reload off. You turn them on under Settings > Usage on claude.ai; inside Claude Code, /usage-credits opens that page in your browser, and the command is not available with API-key sign-in. Under the Help Center's current terms, credits bought on or after December 7, 2026 expire 12 months after purchase, credits bought before that date do not expire, and purchases are non-refundable.

Monthly API credits are a separate perk. Max includes $100 a month, Ultra $200, and Team $20 per Standard seat and $100 per Premium seat, pooled and capped at $500. They are meant for the Claude Platform: the Messages API, the Agent SDK and Managed Agents. They do not cover interactive Claude Code in the terminal, IDE, desktop app or web, and they cannot pay for usage credits. One exception: claude -p run with an API key from the linked Console organization is billed as Agent SDK usage, so the credits cover it. You can claim them after seven days on the plan by linking one Console organization, and unused credits expire at the end of each billing cycle.

Can you get Claude Code cheaper?

Only in a few ways. The Help Center says there are no standard discounts on individual plans and that support cannot issue coupons; promotions are announced through official channels. Annual billing brings Pro down to $17 a month. Eligible nonprofits get discounted Team and Enterprise pricing, and claude.com says Enterprise and API pricing can be customized with volume and committed-spend discounts through sales, at rates it does not publish. Anthropic also gives eligible plans occasional free limit resets, which restore a five-hour or weekly limit rather than lowering the price.

You can also cut what you use rather than what you pay: clear the conversation between tasks, keep Opus or Fable for hard problems and Sonnet for routine edits, and give subagents a smaller model. The cost docs list these habits, and our usage limits guide goes through them in detail.

Frequently asked questions

  • Is Claude Code free? No. The free Claude plan does not include it. The cheapest route is Pro at $20 a month, or $17 a month billed annually, or a Console account that pays per token.
  • Does Claude Code share limits with the Claude app? Yes. On Pro, Max and Ultra, Claude Code in the terminal and IDE and your chats on web, desktop and mobile all draw from the same usage limits.
  • Can I spend Max's $100 of monthly API credits in Claude Code? Not in interactive sessions. The credits cover the Claude API, Agent SDK and Managed Agents, and claude -p only when it runs with an API key from the linked Console organization.
  • Is Fable 5.2 included on Pro? Only through usage credits, billed at API rates. Max, Ultra and Team Premium seats include Fable up to half of the weekly limit.
  • Why am I seeing API charges when I have a subscription? Most often an ANTHROPIC_API_KEY variable is set in your shell, and Claude Code uses it instead of your plan. Unset it, then run /status to confirm the active login.
  • Can I pay for Max or Ultra annually? Not at the moment. Both are monthly-only subscriptions; annual billing is available for Pro and Team.

Source: Claude plans and pricing (claude.com), Claude API pricing (Claude Platform docs), Manage costs effectively (Claude Code docs), Model configuration (Claude Code docs), What is the Pro plan? (Claude Help Center), What is the Max plan? (Claude Help Center), What is the Ultra plan? (Claude Help Center), Use Claude Code with your Pro, Max, or Ultra plan (Claude Help Center), Manage usage credits for paid Claude plans (Claude Help Center), Monthly API credits for Max, Ultra, and Team plans (Claude Help Center), Advanced setup: authenticate (Claude Code docs), Use Claude Code with your Team or Enterprise plan (Claude Help Center), Models, usage and limits in Claude Code (Claude Help Center), Authentication (Claude Code docs), What is a limit reset? (Claude Help Center)