Claude 4.5 Haiku

Reasoning

Creator

Anthropic

Released

2025-10-15

Intelligence

37.1

Artificial Analysis Index

Coding

32.6

Artificial Analysis Index

In $/1M

$1.25

input tokens

Out $/1M

$5.00

output tokens

Blended $/1M

$2.19

3:1 blended

Speed

179

tokens / sec

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
67
Humanity’s Last Exam
10
MMLU-Pro
76
SciCode
43
LiveCodeBench
62
MATH-500
AIME 2025
84
τ²-Bench (agentic)
55
Terminal-Bench Hard
27
IFBench
54
Via Artificial Analysis