All models
Anthropic

Claude Sonnet 4.6

standard

Anthropic

Anthropic's best model for coding and everyday tasks. Exceptional balance of speed, intelligence, and cost — the workhorse of the Claude family.

Official page
best-valuecodingeveryday
Context
1M
Input / 1M
$3.00
Output / 1M
$15.00
Knowledge cutoff
2025-04

Capabilities

Code5/5
Reasoning4/5
Multilingual4/5
Safety5/5
Speed4/5

Modalities & Features

Vision (image input) Yes
Audio No
Video No
Function / tool calling Yes
JSON mode Yes
Extended thinking Yes

Context & Output

Context window1M
Max output tokens64K
Knowledge cutoff2025-04
Release date2025-07

API & Access

Open source No
Streaming Yes
Batch API Yes
Prompt caching Yes
Fine-tuning No

Benchmarks

MMLUGeneral knowledge
89.5%
GPQA DiamondHard science Q&A
68.3%
MATHMath problem solving
86.4%
AIME 2025Advanced math
N/A
GSM8KGrade-school math
95.6%
HumanEvalCode generation
90.8%
LiveCodeBenchLive coding problems
N/A
SWE-bench VerifiedReal-world software
55.3%
HellaSwagCommonsense inference
N/A

Source: provider technical reports & independent evaluations. N/A = not yet published for this model.

Pricing (per 1M tokens)

Input$3.00
Output$15.00
Consumer plan
Claude Pro
$20/mo
Prices shown are list prices and may vary by provider plan. Always check official documentation for the latest pricing.