Claude API (Anthropic API)
Anthropic's metered API for Claude Haiku, Sonnet, Opus and Fable
Overview
The Claude API, long called the Anthropic API, is Anthropic's usage-billed developer access to its Claude models, currently Haiku 5.5, Sonnet 5.5, Opus 5.5 and Fable 5.1, all with a 1M-token context window. It adds batch processing at half price, prompt caching, tool use, web search and fetch, code execution and computer use.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Real-time, latency-sensitive product experiences
- Sub-agents inside larger multi-model systems
- Cache reads $0.01/MTok on prompts up to 100K
- Everyday coding and developer loops
- Customer-facing agents and multi-tool workflows at scale
- Cache reads $0.10/MTok
- Long-horizon coding and large migrations
- Multi-step agents across enterprise tools and data
- Fast mode at $8 / $40 per MTok (research preview)
- Multi-day autonomous projects
- Expert-level work for frontier research
- Cache reads $0.25/MTok
- Opus 5.5 at $2 / $10 per MTok
- Haiku 5.5 from $0.05 / $0.25 per MTok
- Stacks with prompt caching discounts
- Custom rate limits
- Volume discounts
- Dedicated support
- Custom terms
API pricing (pay as you go)
| Model | Input | Cached input | Output |
|---|---|---|---|
| Claude Fable 5.1 Cache read is 0.025x base on this model. | $10 | $0.25 | $50 |
| Claude Opus 5 | $5 | $0.50 | $25 |
| Claude Opus 4.8 | $5 | $0.50 | $25 |
| Claude Sonnet 5 Introductory $2/$10 is now the standard price; planned 1 Sep 2026 rise to $3/$15 cancelled. | $2 | $0.20 | $10 |
| Claude Sonnet 4.6 | $3 | $0.30 | $15 |
| Claude Haiku 4.5 | $1 | $0.10 | $5 |
Use cases
What you can produce with Claude API (Anthropic API)
- Messages API access to Claude Haiku 5.5, Sonnet 5.5, Opus 5.5 and Fable 5.1, each with a 1M-token context window
- Asynchronous batch processing at half the standard token price
- Prompt caching with cache reads at 2.5% to 10% of the input price, depending on the model
- Tool use with server-side web search ($10 per 1,000 searches), web fetch and code execution
- Computer use and browser use toolsets for operating apps that have no API
- Claude Managed Agents sessions billed at token rates plus $0.08 per session-hour
- The same models through Amazon Bedrock, Google Cloud, Microsoft Foundry and Claude Platform on AWS
The Claude API offers four current models with a 1M-token context window across the range, from Haiku 5.5 at $0.10 per million input tokens to Fable 5.1 at $10, plus published discounts that matter in practice: half price for batch jobs and cache reads at 2.5% to 10% of the input price. Tool use, web search and fetch, code execution, computer use and managed agents sit on the same platform, and the models are also sold through Amazon Bedrock, Google Cloud and Microsoft Foundry.
It is a raw API, so teams build their own product around it, and pricing has several modifiers (prompt-length bands on Haiku 5.5, a 1.1x US-only option, fast mode) worth modelling before committing. A strong fit for developers building agents and coding tools who want direct model access.
What we found
The Claude API, long known as the Anthropic API, is Anthropic's pay-as-you-go developer access to its Claude models. It opened beyond a closed alpha on 14 March 2023, when Anthropic introduced Claude through a chat interface and an API in its developer console. The consumer Claude apps and the computer-use capability have separate entries in this directory; this one covers the metered API and the Claude Platform around it.
The current lineup, as listed on Anthropic's pricing page on 14 October 2026, has four models, all with a 1M-token context window. Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100,000 tokens, rising to $0.50 and $2.50 above that. Claude Sonnet 5.5 is $2 and $10, Claude Opus 5.5, which Anthropic suggests as the default for most workloads, is $4 and $20, and Claude Fable 5.1 is $10 and $50. Older models stay available, including Opus 4.5 to Opus 5 at $5 and $25, and Sonnet 5 at $2 and $10, an introductory price Anthropic has since made permanent.
Discounts are where costs move most. The Batch API halves input and output prices for asynchronous jobs. Reading from the prompt cache costs a tenth of the input price on most models, a twentieth on Opus 5.5 and Sonnet 5.5, and a fortieth on Fable 5.1. US-only inference costs 1.1 times standard rates, and a fast mode for Opus 5.5 costs double. Web search adds $10 per 1,000 searches; web fetch adds nothing beyond tokens.
Anthropic's April 2026 postmortem on a spell of complaints about Claude Code's quality traced them to three changes in Claude Code, the Agent SDK and Cowork, and stated that the API was not impacted. Review-site ratings could not be read for this update and are not quoted. Billing is monthly, in US dollars, with payment by card or invoice, usage tracked in the Claude Console, and volume or enterprise terms negotiated with sales.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Claude API (Anthropic API)'s official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Claude API (Anthropic API) unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Claude API (Anthropic API) directly →
Spotted something out of date? Suggest an update →
Related reading
AI & ML
Anthropic Pledges 5% of Its Shares to a New Public Benefit Fund, and Musk Says Give Them t...
AI & ML
Claude Traces the Mysterious NIST P-256 Curve Seeds to a 1997 Clock-Seeded Random-Number G...
Business
Accenture Disputes Reports of a $29 Monthly Claude Cap. The Change Is Who Pays Above It
Alternatives to Claude API (Anthropic API)
More in Code & Dev Tools
