whichdevtool?
Menu
StacksDirectory Pricing changes Glossary
VPS and cloud servers App hosting and PaaS Managed databases Transactional email APIs Email marketing and newsletters Domain registrars CDN, DNS and edge security Monitoring, uptime and error tracking Auth and identity Frontend and full-stack frameworks Backend frameworks UI component libraries AI model APIs File and object storage Payments and billing Background jobs, queues and workflows ORMs and database toolkits

AI model API (model maker) · by Anthropic

Claude API

Anthropic's pay-per-token API for the Claude Fable, Opus, Sonnet and Haiku models, with batch and prompt-caching discounts and hard monthly spend caps per tier.

Visit anthropic.com ↗ Alternatives Trial
Claude API homepage
§1

Overview

The Claude API is Anthropic's direct API for its Claude models. The current line-up on the pricing page runs from Claude Haiku 4.5 at $1 in and $5 out per million tokens to Claude Fable 5.1 at $10 in and $50 out, with Sonnet 5.5 at $2 / $10 and Opus 5.5 at $4 / $20. Claude 4.6 and later models include a 1M-token context window at standard per-token rates.

Two discounts matter most. The Batch API halves input and output prices for asynchronous jobs, and prompt caching bills repeated prompt prefixes at a tenth of the input price (less on Fable 5.1 and Opus 5.5). The two stack. Cached reads also do not count toward input-token rate limits on most models.

Spend is capped per tier: $500 a month on Start, $1,000 on Build and $200,000 on Scale. When an organization reaches its cap, requests return HTTP 429 until the first of the next month unless a higher limit is granted, and you can set a lower limit yourself. New or low-history organizations may begin in an Evaluation tier with lower limits. Anthropic retires older models on a published schedule with at least 60 days' notice.

§2

Pricing and plans

Plan Price What you get
Claude Fable 5.1 $10 in / $50 out per 1M Cache hits $0.25 per 1M
Claude Opus 5.5 $4 in / $20 out per 1M Cache hits $0.20 per 1M; fast mode $8 / $40
Claude Sonnet 5.5 $2 in / $10 out per 1M Cache hits $0.20 per 1M
Claude Haiku 4.5 $1 in / $5 out per 1M Cache hits $0.10 per 1M
Batch API 50% off e.g. Sonnet 5.5 at $1 in / $5 out; stacks with caching
Start tier $500/mo spend cap Build tier $1,000/mo, Scale tier $200,000/mo, Custom tier by agreement

Prices per 1M tokens in USD, read 2026-09-29 on the Claude pricing docs. The FAQ says new users receive a small amount of free credits; no permanent free API tier is published. Prompt cache writes cost 1.25x base input (5-minute) or 2x (1-hour); reads cost 0.1x on most models. US-only inference (inference_geo: "us") adds a 1.1x multiplier on Claude 4.6 and later. Claude 4.7 and later models use a newer tokenizer that produces about 30% more tokens for the same text. Web search costs $10 per 1,000 searches; code execution includes 1,550 free hours per organization per month, then $0.05 per hour. Claude on Bedrock and Google Cloud is priced by those clouds; regional endpoints there carry a 10% premium.

Source: platform.claude.com/docs/en/about-claude/pricing · prices as published, excluding tax.

§3

What it does

Batch API
50% off input and output for asynchronous requests; up to 100,000 requests per batch.
Prompt caching
5-minute or 1-hour cache; hits cost 0.1x base input on most models and do not count toward input rate limits.
Spend caps
Each tier has a monthly cap and you can set a lower organization or workspace limit in the Console.
1M-token context
Claude 4.6 and later models bill a 900K-token request at the same per-token rate as a short one.
Server tools
Web search ($10 per 1,000), web fetch (no extra charge) and code execution (1,550 free hours per month).
Managed Agents
Hosted agent sessions billed at token rates plus $0.08 per session-hour of running time.
§4

Strengths and limitations

Strengths

  • Hard monthly spend caps per tier stop surprise bills; usage pauses instead of accruing.
  • Batch and caching discounts stack, which cuts the cost of large repeated-context jobs sharply.
  • Long context without a long-context surcharge on Claude 4.6 and later models.
  • Published deprecation schedule with at least 60 days' notice before a model is retired.

Limitations

  • No permanent free tier; only a small amount of free credit for new users.
  • Spend caps ($500/month on Start) can stop a growing product until a higher limit is approved.
  • Claude 4.7 and later models produce about 30% more tokens for the same text than earlier models, so per-token prices understate the change.
  • Only Anthropic's own models are available; no open-weight models.
  • First-party data residency option is US-only and costs 1.1x.
§5

Who it suits

Good for

  • Coding agents and tool-using assistants built on Claude
  • Document-heavy apps that reuse long prompts and benefit from caching
  • Teams that want strict monthly spend ceilings

Look elsewhere if

  • Projects that need a free allowance to run indefinitely
  • Teams that need EU-only processing on the first-party API (use Claude through a cloud region instead)
§6

Pricing history

  • Claude Sonnet 5 keeps $2/$10 pricing

    Claude Sonnet 5's launch pricing of $2 in / $10 out per 1M tokens, announced as introductory through 2026-08-31, became the standard price; the increase to $3 / $15 scheduled for 2026-09-01 was cancelled. The announcement date is not stated. Source ↗

  • Claude Opus 4.1 deprecated

    Announced 2026-06-05: claude-opus-4-1-20250805 was retired on the Claude API on 2026-08-05, replaced by Claude Opus 4.8. Source ↗

  • Claude Sonnet 4 and Opus 4 deprecated

    Announced 2026-04-14: claude-sonnet-4-20250514 and claude-opus-4-20250514 were retired on the Claude API on 2026-06-15. Source ↗

  • Claude Haiku 3.5 deprecated

    Announced 2025-12-19: claude-3-5-haiku-20241022 was retired on the Claude API on 2026-02-19, replaced by Claude Haiku 4.5. It remains on Bedrock and Google Cloud. Source ↗

  • Claude 2, 2.1 and Sonnet 3 deprecated

    Announced 2025-01-21: claude-2.0, claude-2.1 and claude-3-sonnet-20240229 were retired on 2025-07-21. Source ↗

§7

Alternatives to Claude API

Tool From Free option Stages
Fireworks AI Inference provider (open-weight models)
Usage-based Trial
Mistral API (Mistral Studio) AI model API (model maker)
Usage-based Free tier
OpenAI API AI model API (model maker)
Usage-based Paid
Together AI Inference provider (open-weight models)
Usage-based Paid
OpenRouter Model router (unified API)
— Free tier
Cerebras Inference Inference provider (open-weight models)
Usage-based Trial

All AI model APIs →

§8

Questions

Is there a free tier for the Claude API?

No permanent free tier. Anthropic's pricing FAQ says new users get a small amount of free credits to test the API.

How much does Claude Sonnet 5.5 cost per million tokens?

On 2026-09-29 it was $2 per 1M input tokens and $10 per 1M output tokens, or $1 and $5 through the Batch API.

What happens when I hit the Claude API spend limit?

Requests return HTTP 429 with error code enforced_spend_limit_reached until 00:00 UTC on the first of the next month, unless you get a higher limit.

Sources read for this page