> Claude Sonnet 4 is a closed AI model from Anthropic, released 22 May 2025. Index score 135 (GPT-4 = 100), rank 104 of 229 ranked models. API price $3.0 in / $15 out per million tokens, 200K context.

[Models](https://themodelindex.org/models/) /[Anthropic](https://themodelindex.org/labs/anthropic/)

# Claude Sonnet 4

Anthropic

Released 22 May 2025

Closed

Flagship

Date verified

[Compare](https://themodelindex.org/compare/claude-sonnet-4-vs-deepseek-r1-0528/) [Announcement](https://www.anthropic.com/news/claude-4)

135123–146

Index score

#104 of 229 models

4/ 8

Core tests taken

3 of 6 areas · plus 19 anchor tests

−12

Versus the frontier at release

best then: o3

Replaced by [Claude Sonnet 4.5](https://themodelindex.org/models/claude-sonnet-4-5/)

$3.0/ $15

Price per million tokens

input / output, list price

200K

Context window

tokens it reads at once

## Where it sits · on the frontier

_Chart: Model Index score of every scored model since late 2022, with the best closed and best open model over time_

o1 reached this score 6 months earlier.

## Test results · by area · tick = best by any model

Show 14 anchor tests

Reasoning & knowledge

evidence

GPQA Diamond

78%

Humanity's Last Exam

8%

Maths

evidence

FrontierMath T1-3

no result

Coding

evidence

SWE-bench Verified

no result

Terminal-Bench

no result

Agents

evidence

APEX-Agents

no result

Novel problems

evidence

ARC-AGI-2

6%

Long tasks

evidence

METR time horizon

1.2 h

Independent results from the Epoch AI benchmarking hub, best reasoning setting. The core tests are shown here; anchor tests also inform the score and can be shown above.

## What others say · rank on each leaderboard

The Model Index

independent tests, our method

#104

of 219 · 135

Epoch Capabilities Index

overall capability from many benchmarks

#98

of 239 · 141.7

Further left is better. Leaderboards measure different things: LMArena is people voting blind between two answers, the others are test scores. Click any row for the source.

## Best at · top three among current models

Replaced by Claude Sonnet 4.5; it no longer leads any category among current models.

Where this model places in the top three of today's models, on category leaderboards and on our core tests.

## Where to use it · prices checked 1 Oct 2026

#### From the lab

Claude

app · subscription

↗

Claude API

API

↗

#### API providers · 1 · per million tokens, in / out

A

Amazon Bedrock

$3.0 / $15

cheapest

Provider prices come from OpenRouter's live listing; going direct to a provider can differ. Subscriptions are the lab's own apps.

## Facts and sources

Released

22 May 2025

Announcement

[anthropic.com ↗](https://www.anthropic.com/news/claude-4)

Release entry

Claude Opus 4 + Sonnet 4

Scored as

Claude Sonnet 4

Artificial Analysis index

–

Output speed

–

Hugging Face

closed weights

## Anthropic releases · around this one

Claude 3.7 Sonnet

Closed

24 FEB 2025

index 134

Claude Opus 4

Closed

22 MAY 2025

index 137

Claude Sonnet 4

Closed

22 MAY 2025

index 135

Claude Opus 4.1

Closed

5 AUG 2025

index 140

Claude Sonnet 4.5

Closed

29 SEP 2025

index 146

---
Source: https://themodelindex.org/models/claude-sonnet-4/ · The Model Index · data as of 2026-10-01