Claude 4.5 Haiku on GPQA Diamond

BENCHMARK RECORD: BR-002H45-002

Report record

Canonical URL: https://www.benchmarkregistry.org/records/BR-002H45-002

Benchmark scores may depend on benchmark version, evaluation configuration, model snapshot, and evaluator. Records sharing a benchmark name are not necessarily directly comparable.

Record

Record Identifier
BR-002H45-002
Record Status
ACTIVE
Created At
2026-07-13T06:16:56.525Z
Updated At
2026-07-13T06:16:56.525Z
Superseded By
Not applicable

Model

Official Model Name
Claude 4.5 Haiku
Model Identifier
ANTHR-H45
Organization
Anthropic
Model Family
Claude
Model Number
45
Tier
H
Model Snapshot
Unknown
Snapshot Label
Unknown
Provider Model Identifier
Unknown
Snapshot Date
Unknown

Benchmark

Benchmark Family
GPQA
Benchmark Version or Variant
Diamond
Benchmark Version Reference
gpqa/diamond
Benchmark Organization
New York University
Benchmark Status
ACTIVE

Result

Metric
Accuracy
Score
67.2%
Numeric Score
67.2
Evaluation Date
Unknown
Reported Date
2026-07-13

Evaluation Context

Configuration Reference
CFG-000007
Shots
Unknown
Reasoning Mode
reasoning
Reasoning Effort
Unknown
Pass Count
Unknown
Agent Scaffold
Unknown
Evaluation Harness
Unknown
Temperature
Unknown
Top P
Unknown
Max Output Tokens
Unknown
System Prompt Description
Unknown
Additional Configuration
None reported

Source and Reporting

Primary Source
Artificial Analysis Model Leaderboard
Source Type
LEADERBOARD
Publisher
Artificial Analysis
Source Published Date
Unknown
Source Accessed At
2026-07-13T04:53:32.160Z
Report Type
INDEPENDENT
Evaluator
Artificial Analysis
Evaluator Type
INDEPENDENT_ORGANIZATION

Provenance

DateEventSourceDetails
2026-07-13T06:16:56.525ZRecord created manuallyArtificial Analysis Model Leaderboard

Record Status

This record is ACTIVE. Registry identifiers remain reserved when record status changes.