Claude Sonnet 3.7 - In-Depth Overview

Anthropic · Claude 3

outdated

Model ID: claude-3-7-sonnet-20250219

Get all the details on Claude Sonnet 3.7, an AI model from Anthropic. This page covers its token limits, pricing structure, key capabilities such as extended_thinking, vision, prompt_caching, available API code samples, and performance strengths.

Key Metrics

Input Limit

200K tokens

Output Limit

64K tokens

Input Cost

$3.00/1M

Output Cost

$15.00/1M

Sample API Code

import anthropic;client = anthropic.Anthropic(api_key="my_api_key",);message = client.messages.create(model="claude-3-7-sonnet-20250219",max_tokens=1024,messages=[{"role": "user", "content": "Hello, Claude!"}]);print(message.content)

Required Libraries

anthropic

@anthropic-ai/sdk

Benchmarks

Benchmark	Score	Source	Notes
lmarena text	1287	lmarena.ai	Rank 34
lmarena webdev	1357	lmarena.ai	Rank 2
lmarena vision	1208	lmarena.ai	Rank 11
livebench global average	64.62	livebench.ai	-
livebench reasoning average	49.11	livebench.ai	-
livebench coding average	74.28	livebench.ai	-
livebench mathematics average	64.65	livebench.ai	-
livebench data analysis average	59.96	livebench.ai	-
livebench language average	63.19	livebench.ai	-
livebench if average	76.49	livebench.ai	-
gpqa	68%	Vellum	-
swe bench	62.3%	Vellum	-
aime 2024	23.3%	Vellum	-
math 500	82.2%	Vellum	-
bfcl	58.3%	Vellum	-
alder polyglot	60.4%	Vellum	-

Notes

High-performance model with early extended thinking. Can increase maximum output token length to 128k tokens by including the beta header output-128k-2025-02-19 in API requests. Streaming Messages API is strongly suggested for longer outputs to avoid timeouts. Model aliases are for experimentation; specific model versions are recommended for production. Models with the same snapshot date are identical across platforms and do not change.