readable AI benchmarks

Anthropic

Claude 4.5 Haiku (Reasoning)

reasoning · closed weights · Oct 15, 2025

Quality Score34.4
Value Score24.7
Reliability47.8
Cache Discount90%

Model specification

Reasoning
reasoning
Input modalities
text, image
Output modalities
text
Context window
200k tokens
Weights
closed weights

Token prices USD per 1M tokens

Input
$1.00
Output
$5.00
Cached Input
$0.100
Cached Output
$1.25
Cost per task
$0.217

Capability

Intelligence
29.9
Coding
43.9
Agentic
16.5
Omniscience
-4.4
Correct
18.0%
Blended price
$0.770

Answer outcomes

Correct 18.0%Incorrect 22.4%Abstained 59.6%

Artificial Analysis benchmarks

GDPval-AA v2
21%
τ³-Banking
9%
Terminal-Bench v2.1
44%
SciCode
43%
Humanity’s Last Exam
10%
GPQA Diamond
67%
CritPt
0%
AA-Omniscience
48%
AA-LCR
74%

Similar models