Inference
EN
Anthropic

Claude 3.7 Sonnet Thinking

Claude 3.7 Sonnet Thinking is a reasoning model from Anthropic in the Claude 3.7 Sonnet family. 15 benchmarks count toward its index, in 7 categories.

  • Closed weights
  • Reasoning
  • Input: text
  • Released 2025-02
RankedEnough results for an index.
IndexAll categories, out of 100.41.4±4.7Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.200KMedian 262K
EloLMArena rating and rank.—Median 1422

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.1 counted39.1—
CodingCode writing and repair.1 counted39.5—
ReasoningLogic problems and puzzles.1 counted38.6—
MultimodalTasks with images and text.2 counted48.169 / 84
KnowledgeFacts and expert knowledge.5 counted44.1172 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.2 counted42.542 / 63
MathMath problems.3 counted36.6132 / 168

Results

15 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Instruction-Following EvalInstruction93.2%15 / 70
MATH 500Math91.6% ±1.221 / 59
MMMLUKnowledge86.1%7 / 15
MATH-500 Problem SetMath96.2%13 / 27
Massive Multi-discipline Multimodal UnderstandingMultimodal75.0%14 / 28
Graduate-Level Google-Proof Q&AKnowledge78.2%72 / 141
Artificial Analysis Omniscience AccuracyKnowledge28.0%87 / 164
Artificial Analysis GPQA DiamondKnowledge77.2%107 / 199
Artificial Analysis IFBenchInstruction48.3%82 / 146
MMLU ProKnowledge82.7% ±0.476 / 135
Critical Physics TasksReasoning0.9%124 / 214
GPQA Diamond · Vals AIKnowledge75.3% ±2.281 / 135
MMMU ProMultimodal75.1% ±1.056 / 92
τ²-Bench Tool-Agent-User EvaluationAgentic54.7%91 / 149
Artificial Analysis Humanity's Last ExamKnowledge9.7%148 / 222
AIMEMath44.6% ±1.764 / 95
LiveCodeBenchCoding60.4% ±1.1106 / 142
American Invitational Mathematics Examination 2024Math61.3%21 / 27
MGSMMultilingual93.0% ±1.611 / 74
TaxEval v2Finance74.0% ±0.942 / 143
MortgageTaxFinance66.9% ±1.831 / 97
LegalBenchLegal82.5% ±0.569 / 149
MedQAHealthcare90.2% ±0.547 / 95
CorpFin v2Finance60.4% ±1.068 / 131

Sources

BenchLMCC BY-NC 4.0Vals AINo licence stated

History

5 releases
Index42.5 → 41.4
23 Oct 202612 Nov 2026

Questions

How good is Claude 3.7 Sonnet Thinking?
Its index is 41.4 out of 100. That is place 204 of 283 ranked models.
What is Claude 3.7 Sonnet Thinking best at?
Knowledge: 44.1 out of 100, place 172 of 268.
What is the context window of Claude 3.7 Sonnet Thinking?
200K tokens.
Is Claude 3.7 Sonnet Thinking open source?
No. Its weights are not published.
When was Claude 3.7 Sonnet Thinking released?
February 2025.

Badge

Claude 3.7 Sonnet Thinking on Inference360: place 204 of 283, index 41.4
HTML
<a href="https://inference360.ai/models/claude-3-7-sonnet-thinking"><img src="https://inference360.ai/badges/claude-3-7-sonnet-thinking.svg" alt="Claude 3.7 Sonnet Thinking on Inference360: place 204 of 283, index 41.4" height="20"></a>
Markdown
[![Claude 3.7 Sonnet Thinking on Inference360: place 204 of 283, index 41.4](https://inference360.ai/badges/claude-3-7-sonnet-thinking.svg)](https://inference360.ai/models/claude-3-7-sonnet-thinking)

Compare with

LFM3-1.7B41.5Granite 4.2 30B41.6Grok Code Fast 141.2Command A+40.2Claude 3.7 Sonnet37.7

More from Anthropic

All models
Claude 3.7 Sonnet37.7Claude 4 Opus Thinking47.0Claude 4 Sonnet39.5Claude 4 OpusUnrankedClaude Sonnet 443.0