✓ ReviewedLast updated October 31, 2026 by Florian Schröder
Claude costs $0 on the Free plan, $20 per month on Pro ($17 per month with annual billing), $100 on Max and $200 on Ultra. Teams pay from $20 per seat, and the API bills per token, from $0.10 per million input tokens on Haiku 5.5 to $10 on Fable 5.1.
Part of our Claude AI Guide. All prices on this page were last verified on October 30, 2026 against Anthropic’s own pricing pages and help center. They are US prices for web subscriptions, before tax. Prices in the mobile apps may differ.
$20 per month, or $17 per month billed annually ($200 up front)
1x, every 5 hours and every week
Yes
Paid with usage credits
Max
$100 per month
5x Pro, every 5 hours and every week
Yes
Up to half of the weekly limit
Ultra
$200 per month
10x Pro per week, no 5-hour limit on verified accounts
Yes
Up to half of the weekly limit
Team, standard seat
$25 per seat per month, or $20 billed annually
More than Pro
Yes
Paid with usage credits
Team, premium seat
$125 per seat per month, or $100 billed annually
5x a standard seat
Yes
Up to half of the weekly limit
Enterprise (self-serve)
$20 per seat per month billed annually (20-seat minimum), plus usage at API rates
The seat includes no usage, cost scales with model and task
Yes, usage at API rates
Yes, at API rates
Team is for 2 to 150 seats, and seat types can be mixed. Sales-assisted Enterprise requires at least 50 seats. Anthropic also offers an Education plan for institutions, without a public price on the pricing page.
A note on older guides: if you read about a “Max 20x” tier at $200 per month, that is the same plan. Anthropic renamed Max 20x to Ultra, at the same price. Its pricing page describes Ultra as 10x Pro usage per week.
Is Claude free?
Yes, Claude has a free plan, but it does not include Claude Code or the Opus models. Free covers chat on web, desktop and mobile with the Sonnet and Haiku models, web search, file creation with code execution, memory, connectors, Artifacts, Skills, voice mode and up to five projects.
What you only get from Pro upward: Claude Code, Opus, Research, more than five projects, Claude in Chrome, Claude for Microsoft 365 and the option to use usage credits. Free also has the lowest usage allowance. Anthropic says Pro gives at least 5x more usage per 5-hour session than Free, but publishes no absolute numbers.
How Claude’s usage limits work
Most subscriptions use two limits at once: a 5-hour session limit and a weekly limit. Anthropic does not publish message or token counts for either one and describes Max and Ultra as multiples of Pro instead. Usage-based Enterprise seats have no per-seat limits, because usage is billed at API rates.
5-hour limit: the period starts with your first message and resets in full five hours later. Verified Ultra accounts do not have this limit. Until an Ultra account is verified, a temporary limit applies.
Weekly limit: it applies across all models and resets at a fixed time assigned to your account. Max, Ultra and premium Team seats can spend up to half of it on Fable, which uses the allowance faster than other models.
Shared allowance: usage in claude.ai, Claude Code and Claude Desktop counts against the same limit.
What uses more: the model you choose, long conversations that trigger automatic context management, large files, higher effort levels, and connected tools and connectors.
When you hit a limit: wait for the reset, upgrade (the new limits apply right away), or turn on usage credits on a paid plan so work continues at standard API rates.
You can see your remaining allowance under Settings > Usage, and with /usage in Claude Code.
Claude Code pricing
Claude Code is not free. It is included in Pro, Max, Ultra and Team, and you can also pay per token through a Claude Console account. Enterprise seats give access, with usage billed at API rates on usage-based plans. On a subscription, it draws on the same limits as chat, including when you run it in VS Code, Cursor or JetBrains IDEs.
Two billing details to check:
A set API key overrides your subscription. If the ANTHROPIC_API_KEY environment variable is set and you approve the key, Claude Code uses it, and API charges apply instead of your plan’s included usage. In non-interactive mode (-p), the key is always used when present.
The API credits on Max and Ultra do not cover interactive Claude Code. The $100 and $200 monthly credits are for building on the Claude Platform. They do not apply to Claude Code in the terminal, IDE, desktop or web, or to the Claude apps.
This is our editorial assessment, based on how the limits are structured. Because Anthropic does not publish absolute limits, start one tier lower than you think you need. Upgrades apply immediately.
If you…
Start with
Why
Want to try Claude or use it occasionally
Free
Sonnet, Artifacts and web search are included at no cost.
Write, research or analyze documents every day
Pro
Adds Opus, Research, Claude Code and more projects for $20.
Code with Claude Code most days and keep hitting Pro limits
Max
5x the allowance on both the 5-hour and the weekly limit.
Run long agent sessions that the 5-hour limit interrupts
Ultra
No 5-hour limit on verified accounts, 10x Pro per week.
Need central billing and SSO for 2 to 150 people
Team
Mix standard and premium seats, no model training on your content by default.
Need SCIM, audit logs and custom data retention
Enterprise
Seat fee plus usage at API rates, with admin spend limits.
Are building a product or automating a pipeline
API
Pay per token. Rate limits and spend caps depend on your API tier.
On Free, Pro, Max and Ultra, model training is an opt-out setting. Team and Enterprise list no model training by default.
The API is billed separately from subscriptions, per million tokens (MTok). These are the four current models:
Model
Input
Output
Cache write (5 min)
Cache read
Fable 5.1
$10
$50
$12.50
$0.25
Opus 5.5
$4
$20
$5
$0.20
Sonnet 5.5
$2
$10
$2.50
$0.10
Haiku 5.5, prompts up to 100K tokens
$0.10
$0.50
$0.125
$0.01
Haiku 5.5, prompts over 100K tokens
$0.50
$2.50
$0.625
$0.05
All four have a 1M-token context window. Except for Haiku 5.5, a long prompt costs the same per token as a short one. Other price factors:
Batch API: 50% off input and output for asynchronous jobs.
Fast mode on Opus 5.5: $8 input and $40 output per MTok. It cannot be combined with the Batch API.
US-only inference: a 1.1x multiplier on all token prices.
Web search: $10 per 1,000 searches, plus token costs for the results.
Code execution: free with web search or web fetch. Otherwise it is billed by execution time after a free allowance.
Older models can cost more. Opus 4.8 is listed at $5/$25, Sonnet 4.6 at $3/$15 and Haiku 4.5 at $1/$5. Moving from Opus 4.8 to Opus 5.5 lowers list prices by 20%. Comparisons with Sonnet 4.6 and Haiku 4.5 need one correction: Anthropic says models from Claude 4.7 onward use a tokenizer that produces about 30% more tokens for the same text. By our calculation, Sonnet 5.5 therefore saves roughly 10 to 13% on the same text against Sonnet 4.6, not the 33% the list price suggests. The exact difference depends on the content. Prompt caching details are in our prompt caching guide.
What the API costs in practice: three worked examples
These are our own calculations with the list prices above, not measurements from a production system. Real costs depend on your prompts and on how long the model’s answers run.
1. Support assistant: 10,000 conversations per month
Assume 2,000 input and 300 output tokens per conversation, which adds up to 20 MTok input and 3 MTok output.
Model
Input cost
Output cost
Total per month
Haiku 5.5
$2
$1.50
$3.50
Sonnet 5.5
$40
$30
$70
Opus 5.5
$80
$60
$140
2. Document questions with prompt caching
Take 1,000 requests on Sonnet 5.5 against the same 50,000-token document, each with 500 new input tokens and 500 output tokens.
Without caching: 50.5 MTok input ($101) plus 0.5 MTok output ($5) comes to $106.
With caching, assuming every request after the first is a cache hit: one cache write ($0.13), 999 cache reads ($5.00), new input ($1) and output ($5) come to about $11, roughly 90% less.
3. Overnight batch job: 100,000 short documents
With 3,000 input and 300 output tokens per document, that is 300 MTok input and 30 MTok output. Sonnet 5.5 costs $900 at standard rates and $450 with the Batch API. Haiku 5.5 with the Batch API costs $22.50.
What a subscription fee would buy on the API
This table is a yardstick, not a statement of what a plan includes, because Anthropic does not publish plan limits in tokens. It assumes a mix of three input tokens for each output token, on the Haiku 5.5 price for prompts up to 100K tokens.
Model
$20 buys
$100 buys
$200 buys
Fable 5.1
1 MTok
5 MTok
10 MTok
Opus 5.5
2.5 MTok
12.5 MTok
25 MTok
Sonnet 5.5
5 MTok
25 MTok
50 MTok
Haiku 5.5
100 MTok
500 MTok
1,000 MTok
Anthropic puts 1 MTok at roughly 555,000 words on the current tokenizer.
Frequently asked questions
How much does Claude cost per month?
Free costs $0, Pro $20 ($17 with annual billing), Max $100 and Ultra $200. Team seats cost $25 or $125 per month, or $20 and $100 with annual billing. Prices exclude tax.
What is the difference between Claude Max and Ultra?
Max costs $100 and gives 5x Pro usage on both the 5-hour and the weekly limit. Ultra, formerly called Max 20x, costs $200, gives 10x Pro usage per week and removes the 5-hour limit on verified accounts. It also includes $200 instead of $100 in monthly API credits.
Is Claude Pro worth it?
In our assessment, yes, if you use Claude most days and want Opus, Research or Claude Code. If you only ask occasional questions, start on Free. If you code with Claude Code for hours every day, expect to reach Pro limits and compare Max.
Does a Claude subscription include API access?
Only as credits, and not on Free or Pro. Max includes $100 and Ultra $200 in API credits per month. Team includes $20 per standard seat and $100 per premium seat, pooled and capped at $500 per month. Unused credits expire at the end of each billing cycle. When they run out, API requests stop unless the organization has other credits.
What are usage credits?
Usage credits let paid plans keep working after reaching a limit instead of waiting for the reset. They are billed at standard API rates and are not available on Free. On Pro and standard Team seats, they are also the way to use Fable.
Sources
Prices and plan facts come from Anthropic’s own pages. The calculations and plan recommendations are ours. Prices and plans can change at Anthropic’s discretion, so check them before you buy or budget.
Florian Schröder is a co-founder of AI Rockstars and an expert in AI strategies for marketing. He advises companies on integrating AI tools such as ChatGPT, Claude, and Gemini into their workflows. With over 10 years of experience in digital marketing, he combines technical expertise with practical applications.
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
