Here's What Every Claude Model Costs on Anthropic's API in October 2026

The four current models

Anthropic's models overview names Claude Opus 5.5 as the place to start for most workloads and Claude Fable 5.1 for demanding reasoning and long-horizon agentic work. All four current models have a 1M-token context window and up to 128K output tokens. Prices are quoted in US dollars per million tokens.

Claude Fable 5.1 costs $10 per million input tokens and $50 per million output tokens. Claude Opus 5.5 costs $4 for input and $20 for output, and Claude Sonnet 5.5 costs $2 for input and $10 for output. Claude Haiku 5.5 is the only current model priced by prompt length: a prompt of up to 100,000 tokens costs $0.10 per million input tokens and $0.50 per million output tokens, while a prompt over 100,000 tokens costs $0.50 for input and $2.50 for output.

Prompt caching

Anthropic charges for writing to the cache and gives a discount for reading from it. A five-minute cache write costs 1.25 times the base input price and a one-hour cache write costs twice the base input price. A cache read normally costs 0.1 times the base input price, but Anthropic sets lower read rates on the newest models: 0.025 times on Claude Fable 5.1 and Claude Mythos 5.1, and 0.05 times on Claude Opus 5.5 and Claude Sonnet 5.5.

In dollars, that puts cache writes on Claude Fable 5.1 at $12.50 for five minutes and $20 for one hour, with reads at $0.25. On Claude Opus 5.5 the figures are $5, $8 and $0.20, and on Claude Sonnet 5.5 the figures are $2.50, $4 and $0.10. On Claude Haiku 5.5, prompts up to 100,000 tokens cost $0.125, $0.20 and $0.01, and longer prompts cost $0.625, $1 and $0.05. Anthropic says the caching multipliers stack with other modifiers, including the Batch API discount and data residency, and that at the standard 10% read rate caching pays off after one read for a five-minute write or two reads for a one-hour write.

Batch API

The Batch API processes large volumes of requests asynchronously at a 50% discount on both input and output tokens. That makes Claude Fable 5.1 $5 and $25 per million tokens, Claude Opus 5.5 $2 and $10, and Claude Sonnet 5.5 $1 and $5. Claude Haiku 5.5 costs $0.05 and $0.25 for prompts up to 100,000 tokens and $0.25 and $1.25 for longer ones. Anthropic's FAQ says batch and prompt caching discounts can be combined.

Long context

According to Anthropic, Claude Mythos Preview and every Claude model from 4.6 onward except Haiku 5.5 include the full 1M-token context window at standard pricing, so a 900k-token request is billed at the same per-token rate as a 9k-token request. Claude Haiku 5.5 is the exception. A request whose prompt is over 100,000 tokens pays the higher prices, the prompt length counts every input token including cache reads and writes, and each request is priced on its own, so a request over the threshold pays the higher prices even when part of the prompt is a cache hit.

US-only inference and fast mode

For Claude 4.6 and later models, setting the inference_geo parameter to US-only inference applies a 1.1 times multiplier to every token category, including input, output, cache writes and cache reads, and on Claude Haiku 5.5 it also applies to the higher long-prompt prices. Global routing, the default, uses standard pricing. The same multiplier applies on Claude Platform on AWS and to Microsoft Foundry deployments that use the US Data Zone Standard type. Earlier models don't support the parameter and return a 400 error if it's included.

Fast mode gives faster output at premium prices. On Claude Opus 5.5 it costs $8 per million input tokens and $40 per million output tokens, and on Claude Opus 5 and Claude Opus 4.8, where it's in research preview, it costs $10 and $50. Anthropic says fast mode pricing applies across the full context window, prompt caching and data residency multipliers apply on top of it, and fast mode isn't available with the Batch API. It's offered on the Claude API and, for Claude Opus 5.5, on Microsoft Foundry deployments hosted on Azure, but not on Claude Platform on AWS or partner-operated clouds.

Older models on the price list

Anthropic's pricing page still lists a set of additional models. Claude Mythos 5.1 costs $10 and $50 per million tokens with $0.25 cache reads, and Claude Fable 5 and Claude Mythos 5 cost $10 and $50 with $1 cache reads. Claude Opus 5, Opus 4.8, Opus 4.7, Opus 4.6 and Opus 4.5 cost $5 and $25, while Claude Opus 4.1 and Opus 4 cost $15 and $75. Claude Sonnet 5 costs $2 and $10, and Claude Sonnet 4.6, Sonnet 4.5 and Sonnet 4 cost $3 and $15; the page marks Sonnet 4.5 as deprecated. Claude Haiku 4.5 costs $1 and $5, and Claude Haiku 3.5 costs $0.80 and $4. The Batch API halves each of these prices as well.

Token counts can differ between models. Anthropic says Claude 4.7 and later models and Claude Mythos Preview use a newer tokenizer that produces about 30% more tokens for the same text than earlier models, and suggests counting a request's tokens with the specific model's ID.

Tools and agents

Tool use is billed through ordinary tokens, plus a system prompt the API adds when tools are present. On Claude Opus 5.5, Sonnet 5.5 and Haiku 5.5 that system prompt is 286 tokens when tool choice is auto or none, and on Haiku 5.5 it's 406 tokens when tool choice is set to any or to a specific tool. Web search costs $10 per 1,000 searches plus standard token costs, and a search that errors isn't billed. Web fetch has no charge beyond the tokens of the fetched content. Code execution is free when used together with the current web search or web fetch tools; used on its own, it's billed by execution time with a five-minute minimum, and Anthropic's pricing page lists 50 free hours a day per organization, which the pricing documentation expresses as 1,550 free hours a month; after that it's $0.05 per hour per container.

Claude Managed Agents bills tokens at the standard model rates and adds $0.08 per session-hour of runtime, measured only while a session is running. Anthropic's worked example puts a one-hour Claude Opus 5.5 session with 50,000 input tokens and 15,000 output tokens at $0.58, or $0.428 if 40,000 of the input tokens are cache reads. The Batch API discount doesn't apply to Managed Agents sessions.

Credits, tiers and discounts

Anthropic says new users receive a small amount of free credits to test the API, without giving an amount. Rate limits depend on usage tier, which the pricing page names as Start, Build and Scale. Volume discounts are negotiated case by case, academic and research discounts may be available, and billing is based on actual monthly usage, paid in US dollars. On Amazon Bedrock and Google Cloud, the cloud provider sets and invoices the price, and Anthropic says regional and multi-region endpoints there carry a 10% premium over global endpoints for Claude Sonnet 4.5, Haiku 4.5, Opus 4.5 and later models.