Claude API pricing guide & calculator
Claude API prices range from $0.10 to $10 per million input tokens and $0.50 to $50 per million output tokens, depending on the model. Below: every current model’s price, what common jobs cost per month, and a calculator for your own usage.
Current Claude API prices
| Model | Input | Cached input | Output | Batch in / out |
|---|---|---|---|---|
| Claude Fable 5.1flagship · 1M tokens | $10 | $0.25 | $50 | $5 / $25 |
| Claude Opus 5.5flagship · 1M tokens | $4 | $0.20 | $20 | $2 / $10 |
| Claude Sonnet 5.5balanced · 1M tokens | $2 | $0.10 | $10 | $1 / $5 |
| Claude Haiku 5.5fast | $0.10 | $0.01 | $0.50 | $0.05 / $0.25 |
Per million tokens, standard tier, in US dollars. Checked October 10, 2026 against Anthropic’s official pricing. “Cached input” is the price for text you reuse across requests, like long instructions.
- Claude Fable 5.1: Anthropic’s most capable generally available model.
- Claude Haiku 5.5: Prompts over 100,000 tokens are billed at $0.50 / $2.50.
Estimate your monthly Claude API cost
Pick a job and how many you’ll run a month. Claude’s models are compared with OpenAI and Gemini; tap a provider to add or remove it. How the estimates work.
| Model | Per ticket | Per month | vs. cheapest |
|---|---|---|---|
| Claude Haiku 5.5Anthropic · fast | $0.00042 | $0.85 | 1× |
| GPT-6 LunaOpenAI · fast | $0.00042 | $0.85 | 1× |
| Gemini 3.5 Flash-LiteGoogle · fast | $0.0016 | $3.25 | 3.8× |
| Gemini 3.8 FlashGoogle · balanced | $0.0032 | $6.38 | 7.5× |
| Claude Sonnet 5.5Anthropic · balanced | $0.0085 | $17.00 | 20× |
| GPT-6.1 SolOpenAI · balanced | $0.0085 | $17.00 | 20× |
| Gemini 3.1 Pro (Preview)Google · flagship | $0.0092 | $18.40 | 22× |
| Claude Opus 5.5Anthropic · flagship | $0.0170 | $34.00 | 40× |
| Claude Fable 5.1Anthropic · flagship | $0.0425 | $85.00 | 100× |
| GPT-6 AstraOpenAI · flagship | $0.0425 | $85.00 | 100× |
Published list prices checked October 10, 2026, before tax. Token counts are estimates and vary by provider, since each counts text a little differently. Subscriptions like Claude Pro, ChatGPT Plus, or Google AI Pro (about $20 a month) are for using the chat apps yourself and can’t be called from your own software. Calculated in your browser; nothing is saved.
Which Claude model should you use?
Anthropic sells four current models through the API, and each step down the line is a big step down in price. Sonnet 5.5 is where most business applications should start: at $2 per million input tokens and $10 per million output tokens, it handles writing, summarizing, customer replies, and most analysis well, and it’s a fifth of the price of the top model.
Haiku 5.5 is twenty times cheaper than Sonnet and is built for high-volume, simpler jobs: sorting incoming email, tagging support tickets, pulling fields out of forms, or short product descriptions. For short prompts it’s one of the cheapest capable models from any major provider.
Opus 5.5 and Fable 5.1 are for the hardest work: long reasoning, complex coding, multi-step agents, and anything where a better answer is worth paying for. Fable 5.1 is Anthropic’s most capable generally available model and costs five times as much as Sonnet. A sensible pattern is to build with Sonnet, test whether Haiku is good enough for the simple steps, and reserve Opus or Fable for the steps where quality visibly improves.
What common jobs cost on Claude
Monthly cost at standard prices, with no caching or batch discounts, for typical volumes. The token sizes behind each job are explained in how we calculate.
| Job (per month) | Fable 5.1 | Opus 5.5 | Sonnet 5.5 | Haiku 5.5 |
|---|---|---|---|---|
| 1,000 × draft an email | $28 | $11.2 | $5.6 | $0.28 |
| 2,000 × answer a support ticket | $85 | $34 | $17 | $0.85 |
| 300 × summarize a 10-page document | $31.5 | $12.6 | $6.3 | $0.315 |
| 100 × write a 1,500-word article | $11.2 | $4.48 | $2.24 | $0.112 |
| 5,000 × write a product description | $90 | $36 | $18 | $0.90 |
| 3,000 × chatbot conversation (10 turns) | $660 | $264 | $132 | $6.6 |
| 200 × ai coding or research agent task | $400 | $160 | $80 | $20 |
For a sense of scale: answering 2,000 support tickets a month costs about $17 on Sonnet 5.5 and $0.85 on Haiku 5.5. The expensive row is the chatbot, because every turn of a conversation resends everything said so far, so input tokens pile up quickly.
How Claude compares on price
Claude and OpenAI price their lineups almost identically, tier for tier: Fable 5.1 and GPT-6 Astra are both $10 / $50, Sonnet 5.5 and GPT-6.1 Sol are both $2 / $10, and Haiku 5.5 and GPT-6 Luna are both $0.10 / $0.50. Google’s balanced model, Gemini 3.8 Flash, is the cheapest of the three. DeepSeek’s and Grok’s top models cost less than Sonnet 5.5 too, though more than Gemini Flash. The table compares the big three’s balanced models on 2,000 support tickets a month.
| Balanced model | Input / output per 1M | 2,000 tickets |
|---|---|---|
| Gemini 3.8 FlashGoogle | $0.75 / $3.75 | $6.38 |
| Claude Sonnet 5.5Anthropic | $2 / $10 | $17 |
| GPT-6.1 SolOpenAI | $2 / $10 | $17 |
Discounts: batch, caching, and free use
Batch processing: 50% off. If results can wait (Anthropic processes most batches within an hour, and always within 24 hours), the Batch API halves both input and output prices. Overnight document processing, bulk product descriptions, and weekly report generation are good fits.
Prompt caching: up to 97.5% off repeated input. When you send the same long instructions or reference material with every request, Claude can cache it. Reading from the cache costs a fraction of the normal input price: 0.05× on Opus 5.5 and Sonnet 5.5, 0.025× on Fable 5.1, and 0.1× on other models. Writing to the cache costs a little more than normal input (1.25× for a five-minute cache, 2× for an hour), so caching pays off when the same text is reused several times. For a support bot that sends the same help-center content with every ticket, it can cut the input bill dramatically.
The discounts stack: batch processing applies on top of cache pricing.
Free tier: No free API tier; new Console accounts may get a small trial credit.
What surprises people on the bill
Newer Claude models use more tokens for the same text
Anthropic says Claude 4.7 and later use a newer tokenizer that produces roughly 30% more tokens for the same text than its older models. That means a per-token price comparison with older Claude models, or with other providers, understates Claude’s real cost by up to about that much, depending on the content. Test with your own text before you budget.
Haiku 5.5 gets five times pricier on long prompts
Haiku 5.5 is priced by prompt length: any request with more than 100,000 input tokens is billed at $0.50 / $2.50 instead of $0.10 / $0.50, for the whole request. On our long agent task, Haiku 5.5 costs $20 a month versus $4 for GPT-6 Luna, which only raises its price above 272,000 tokens. Keep Haiku prompts short, or use Sonnet for long-document work.
Long context is otherwise flat-priced
The good news on the bigger models: Fable 5.1, Opus 5.5, and Sonnet 5.5 include the full 1 million token context window at standard pricing, so a 900,000-token request is billed at the same per-token rate as a short one. OpenAI and Google both charge more above a threshold.
Extras are billed separately
Server-side tools like web search are charged per use on top of tokens, and requesting US-only processing adds a 1.1× multiplier to every token price on Claude 4.6 and later models. Opus also offers a faster “fast mode” at premium prices. None of these are included in the calculator.
Claude API vs. Claude Pro
The API and the Claude subscriptions are separate products with separate bills. Claude Pro ($20 a month) is for using Claude yourself in the app, on the web, desktop, and mobile, including Cowork and Claude Code. The API is for building Claude into your own software, website, or automations, and you pay per token. A Pro subscription doesn’t give you API credits, and API credits don’t unlock the Claude app.
To put the numbers side by side: $20 of API usage buys roughly 3,571 email drafts on Sonnet 5.5, or about 71,428 on Haiku 5.5, using our token estimates. If you’re one person drafting emails and documents by hand, Pro is simpler and includes the apps. If you want Claude to answer customers, process documents, or run inside another tool automatically, you need the API. Our guide to Claude subscription pricing covers the plans in detail.
| Anthropic subscription | Price |
|---|---|
| Claude Pro | $20/mo ($17/mo billed annually) |
| Claude Max 5x | $100/mo |
| Claude Max 20x | $200/mo |
Who the Claude API is a good fit for
Claude is a strong choice for writing-heavy work, long documents, and agents that need to follow detailed instructions reliably. The flat-priced 1 million token context window on Sonnet, Opus, and Fable makes it especially good value for jobs that read a lot of text at once, like reviewing contracts or analyzing a large set of files.
It’s less of a bargain for very high-volume, very short tasks where every fraction of a cent matters; there, Haiku competes well, but Gemini Flash-Lite and DeepSeek Flash are worth testing. For anything a customer will see, test quality first: the cheapest model that’s good enough is almost always the right answer. To see how Claude is used day to day outside the API, the free Claude Cowork course walks through real business workflows.
Frequently asked questions
How much does the Claude API cost?
Claude API prices depend on the model. Per million tokens: Fable 5.1 is $10 input / $50 output, Opus 5.5 is $4 / $20, Sonnet 5.5 is $2 / $10, and Haiku 5.5 is $0.10 / $0.50 for prompts up to 100,000 tokens. Batch processing halves those prices, and cached input is much cheaper.
Is the Claude API free?
No. There’s no ongoing free tier for the Claude API; you pay per token through an Anthropic Console account. New accounts sometimes receive a small trial credit. The free Claude.ai plan covers chat in the app only, not the API.
Does Claude Pro include API access?
No. Claude Pro is a subscription for using Claude in the app, including Cowork and Claude Code. API usage is billed separately, per token, through the Anthropic Console.
Which Claude model is cheapest?
Claude Haiku 5.5, at $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100,000 tokens. Above that length it costs $0.50 / $2.50. With the Batch API, it’s half that again.
Is the Claude API cheaper than OpenAI?
At list price, tier for tier, they’re the same: Fable 5.1 matches GPT-6 Astra, Sonnet 5.5 matches GPT-6.1 Sol, and Haiku 5.5 matches GPT-6 Luna. Differences come from caching discounts, long-prompt pricing, and how many tokens each model uses for the same text, so test your own workload.