Inference
EN
Anthropic

Claude Sonnet 4.5 Thinking

Claude Sonnet 4.5 Thinking is a reasoning model from Anthropic in the Claude Sonnet 4.5 family. 11 benchmarks count toward its index, in 6 categories.

  • Closed weights
  • Reasoning
  • Input: text
  • Released 2025-09
RankedEnough results for an index.
IndexAll categories, out of 100.53.8±6.9Median 51.0
CoverageIndex weight that has results.80%
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.2 counted57.189 / 218
CodingCode writing and repair.4 counted50.6134 / 209
ReasoningLogic problems and puzzles.None counted——
MultimodalTasks with images and text.1 counted58.4—
KnowledgeFacts and expert knowledge.2 counted53.3110 / 268
MultilingualTasks in many languages.1 counted60.4—
InstructionTasks with strict prompt rules.None counted——
MathMath problems.1 counted54.0—

Results

11 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Terminal-Bench 1.0Agentic61.3% ±5.42 / 47
EuroEval PolishMultilingual63.8% ±0.310 / 209
EuroEval DutchMultilingual65.3% ±0.410 / 175
EuroEval GermanMultilingual57.1% ±0.514 / 191
EuroEval SpanishMultilingual60.4% ±0.415 / 176
EuroEval SwedishMultilingual66.0% ±0.617 / 195
EuroEval PortugueseMultilingual64.8% ±0.417 / 182
EuroEval FrenchMultilingual66.9% ±0.421 / 178
EuroEval ItalianMultilingual65.1% ±0.321 / 174
MMLU ProKnowledge87.4% ±0.428 / 135
AIMEMath88.2% ±0.730 / 95
IOI v1Coding18.3% ±5.925 / 61
Terminal-Bench 2.0 · Vals AIAgentic41.6% ±5.329 / 66
GPQA Diamond · Vals AIKnowledge81.6% ±2.361 / 135
MMMU ProMultimodal79.3% ±1.048 / 92
LiveCodeBenchCoding73.0% ±1.182 / 142
SWE-benchCoding70.0% ±2.161 / 87
Vibe Code Bench v1.1Coding22.6% ±3.787 / 123
MGSMMultilingual94.3% ±1.44 / 74
MedQAHealthcare94.7% ±0.215 / 95
LegalBenchLegal84.1% ±0.439 / 149
Finance Agent v1.1Finance54.5% ±2.916 / 51
MedScribeHealthcare84.1% ±1.938 / 109
TaxEval v2Finance73.3% ±0.950 / 143
CaseLaw v2Legal62.2% ±1.519 / 53
CorpFin v2Finance62.0% ±1.048 / 131
MedCodeHealthcare44.1% ±2.042 / 106
MortgageTaxFinance64.0% ±1.049 / 97

Sources

BenchLMCC BY-NC 4.0Vals AINo licence statedEuroEvalMIT

History

5 releases
Index55.0 → 53.8
23 Oct 202612 Nov 2026

Questions

How good is Claude Sonnet 4.5 Thinking?
Its index is 53.8 out of 100. That is place 109 of 283 ranked models.
What is Claude Sonnet 4.5 Thinking best at?
Agentic: 57.1 out of 100, place 89 of 218.
What is the context window of Claude Sonnet 4.5 Thinking?
200K tokens.
Is Claude Sonnet 4.5 Thinking open source?
No. Its weights are not published.
When was Claude Sonnet 4.5 Thinking released?
September 2025.

Badge

Claude Sonnet 4.5 Thinking on Inference360: place 109 of 283, index 53.8
HTML
<a href="https://inference360.ai/models/claude-sonnet-4-5-thinking"><img src="https://inference360.ai/badges/claude-sonnet-4-5-thinking.svg" alt="Claude Sonnet 4.5 Thinking on Inference360: place 109 of 283, index 53.8" height="20"></a>
Markdown
[![Claude Sonnet 4.5 Thinking on Inference360: place 109 of 283, index 53.8](https://inference360.ai/badges/claude-sonnet-4-5-thinking.svg)](https://inference360.ai/models/claude-sonnet-4-5-thinking)

Compare with

GPT-5.1-Codex-Max53.8Claude Sonnet 4.547.9Claude Opus 4.147.5

More from Anthropic

All models
Claude Sonnet 4.547.9Claude Haiku 4.5 Thinking43.5Claude Haiku 4.543.2Claude Opus 4.147.5Claude 4.1 Opus Thinking47.0