← All providers

Anthropic API Pricing, Models & Rate Limits (2026)

Anthropic trains the Claude family, built around a Messages API and, more recently, extended-thinking reasoning modes on its top-tier models. Claude Fable 5 is the current flagship, with Opus, Sonnet, and Haiku tiers trading intelligence for cost and speed below it.

Current models
4
Legacy models
7
Price range /M
$2.00–$20.00
Max context
1M
Median tok/s
67
Next retirement

Models and prices

ModelInput /MOutput /MBlended /M
Claude Haiku 4.5$1.00$5.00$2.00
Claude Sonnet 4.6$3.00$15.00$6.00
Claude Opus 4.8$5.00$25.00$10.00
Claude Fable 5$10.00$50.00$20.00
7 legacy Anthropic models
Claude Sonnet 4.5$6.00/M blended
Claude Sonnet 4$6.00/M blended
Claude Opus 4.7$10.00/M blended
Claude Opus 4.6$10.00/M blended
Claude Opus 4.5$10.00/M blended
Claude Opus 4.1$30.00/M blended
Claude Opus 4$30.00/M blended

Speed

Fastest measured Anthropic model is Claude Haiku 4.5 at 148 tokens/sec (260ms TTFT), median across measured Anthropic models is 67 tokens/sec. See the full speed benchmark methodology.

Operational facts

Verified 2026-08-08 against source.

OpenAI-compatiblePartial
API base URLhttps://api.anthropic.com/v1
Auth modelx-api-key header
Prompt cachingYes
Batch discount50%
Free tierNone
Rate-limit modelUsage tiers 1-4, auto-promoted by spend
Data residencyNot documented
Trains on API dataNo
SLA publishedYes
DocsOfficial pricingStatus page

Lifecycle

Anthropic has 7 legacy models still routable. Full dates and successors on the model deprecation tracker.

Switching to and from Anthropic

The closest-priced current model to Claude Fable 5 ($20.00/M) outside Anthropic is GPT-5.6 Sol ($11.25/M, -43.8%). Same request shape through All AI Ask — swap the model id, keep the rest of your code.
The closest-priced current model to Claude Opus 4.8 ($10.00/M) outside Anthropic is GPT-5.6 Sol ($11.25/M, +12.5%). Same request shape through All AI Ask — swap the model id, keep the rest of your code.

Calling Anthropic through All AI Ask

Calling Anthropic directly means adapting client code away from the plain OpenAI request shape — our gateway removes that: every model, including Anthropic's, is called the same way.

curl https://allaiask.com/api/v1/chat \
  -H "Authorization: Bearer $ALLAIASK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "claude-haiku-4-5", "messages": [{"role": "user", "content": "Hello"}]}'

FAQ

Is Anthropic OpenAI-compatible?

Partially. Anthropic publishes an OpenAI-compatible mode for some endpoints, but not a full 1:1 replacement for its native API — check https://docs.anthropic.com before relying on it for every feature you use.

Does Anthropic support prompt caching?

Yes, as of 2026-08-08 — see https://docs.anthropic.com for the current mechanics and discount.

Does Anthropic have a free tier?

No published free tier as of 2026-08-08 — billing starts from the first request.

How much does the Anthropic API cost?

Current Anthropic models range from $2.00 to $20.00 per million blended tokens (3:1 input:output). Full per-model pricing is in the table below.

Where is Anthropic API data hosted?

Not documented as of 2026-08-08 — no published data-residency commitment found for Anthropic.

Try Anthropic for free

Run real prompts against every current Anthropic model, and every other provider on this site, in one workspace.

Try It Free