Claude Sonnet 4.5

Frontier‑level coding and agentic performance. Claude Sonnet 4.5 leads on real‑world coding tasks and shows substantial gains in computer use, reasoning, and math. Designed for long‑horizon, multi‑step work across IDE + terminal workflows, large codebases, and complex agents — a drop‑in upgrade over previous Sonnet versions.

Reasoning
Vision
Native PDF input
Tool Calling
Structured Output

Added Sep 29, 2025

Pricing

Auto routing · per 1M tokens

Input: $3.00
Output: $15.00
Cache read: $0.30

Compare provider prices

Specifications

Context window: 200K
Max output: 64K
Avg output (7d): 289 tokens; Longer than 23% of models

Benchmarks

Artificial Analysis

LMArena

Vectara

Design Arena

Sourced from Artificial Analysis.

Intelligence Index

19.3

Better than 67% of models compared

Agentic work

T²-Bench Telecom (legacy)

Legacy fallback · Conversational AI agents in dual-control scenarios

70.5%

Better than 63% of models compared

Document reasoning

AA-LCR v1.1

Long context reasoning with updated grading

54.0%

Better than 47% of models compared

Reasoning

HLE

Humanity's Last Exam

7.2%

Better than 42% of models compared

IFBench

Instruction-following benchmark

42.7%

Better than 45% of models compared

Coding

Terminal-Bench Hard (legacy)

Legacy fallback · Agentic coding and terminal use

28.8%

Better than 71% of models compared

LiveCodeBench

Contamination-free coding benchmark

59.0%

Better than 65% of models compared

Math

AIME 2025

American Invitational Mathematics Examination 2025

37.0%

Better than 38% of models compared

Knowledge

MMLU-Pro

Professional and academic subject knowledge

86.0%

Better than 94% of models compared

Legacy benchmarks

GPQA Diamond (legacy)

Graduate-level scientific reasoning

72.7%

Better than 52% of models compared

AA-LCR (unversioned / legacy)

Long context reasoning evaluation

54.0%

Better than 46% of models compared

Last updated Oct 14, 2026