Claude 4.5 Sonnet
Anthropic · 2 thinking configurations
Intelligence20.7
Input / 1M$3.00
Output / 1M$15.00
Context1M
Tariff: Anthropic · Source: artificial-analysis · USD per million tokens
Thinking configuration
What does more thinking buy you?
Compare this model’s measured thinking configurations.
Scroll the chart ↔ · tap a point for details
Efficient frontier: best measured score at each cost1 measured configurations · 1 without both values
Selected: thinking thinking · Click a point to switch configuration
Claude Sonnet 4.5 is Anthropic’s most advanced Sonnet model to date, optimized for real-world agents and coding workflows. It delivers state-of-the-art performance on coding benchmarks such as SWE-bench Verified, with...
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
20.7 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
83.4 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
17.8 %SciCodeCode generation for scientific research problems.
45.7 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
55.8 %τ²-BenchTool use and interaction in assistance scenarios. Does not replace an evaluation of your integration.
78.1 %IFBenchVerifiable instruction following. A proxy for precision rather than a benchmark of creative style.
57.3 %MMMU-ProMultimodal understanding of academic problems that include images.
68.7 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
32.9 %Specifications from the catalog
Inputtext, image, file
Outputtext
Tool callingYes
Structured outputYes
Speed40.0 tok/s
First token13.78 s
Maximum output64K
Catalog listing date2025-09-29