modelgrep
A

Anthropic: Claude Sonnet 4

anthropic/claude-sonnet-4

66th smartest of 180ReasoningToolsVision
Use via OpenRouter ↗
Intelligence
33.0
66th of 180
Design Elo
1220
3D
Speed
41
201st fastest
Latency
658ms
first token
Input price
$3.00
267th cheapest
Context
1M
64K max out

How it compares

Smarter than63%
of all ranked models
Faster than35%
of all ranked models
Cheaper than10%
of all ranked models

Overview

Claude Sonnet 4 significantly enhances the capabilities of its predecessor, Sonnet 3.7, excelling in both coding and reasoning tasks with improved precision and controllability. Achieving state-of-the-art performance on SWE-bench (72.7%),...

Benchmarks

independent · via OpenRouter
Artificial Analysis60th percentile
Intelligence Index
33.0
Coding Index
30.6
Agentic Index
39.2
GPQA Diamond
68%
Humanity's Last Exam
4%
SciCode
37%
Tau²-Bench (agentic)
52%
Design Arena · Elo8,313 tournaments
Game Dev
1208
3D
1207
Data Viz
1196
codecategories
1188
Website
1184
UI Component
1183
svg
1135

Providers & pricing (5)

ProviderIn $/MOut $/MUptime
Google$3.00$15.00100%
Amazon Bedrock$3.00$15.00
Amazon Bedrock$3.00$15.0099.9%
Google$3.00$15.00
Google$3.00$15.00

Specifications

Context window1M
Max output64K
Knowledge cutoffJan 2025
Input modalitiesimage, text, file
Output modalitiestext
Prompt caching
Cache read price$0.300/M
ModeratedNo

Claude Sonnet 4 FAQ

How much does Claude Sonnet 4 cost?

Claude Sonnet 4 costs $3.00 per million input tokens and $15.00 per million output tokens via OpenRouter, making it 267th cheapest of 298 paid models.

How smart is Claude Sonnet 4?

Claude Sonnet 4 scores 33.0 on the Artificial Analysis Intelligence Index, ranking 66th of 180 benchmarked models, with a GPQA Diamond score of 68%.

How fast is Claude Sonnet 4?

Claude Sonnet 4 generates around 41 tokens per second with 658ms time-to-first-token (p50), the 201st fastest tracked model.

What is Claude Sonnet 4's context window?

Claude Sonnet 4 supports a 1M-token context window and can output up to 64K tokens. It accepts image, text, file input.

Compare head-to-head