← Claude Sonnet 5 · API · High effort (default)
Evidenceacceptedindependent_lab
GPQA Diamond
Artificial Analysis — GPQA Diamond (Claude Sonnet 5 max)
ev_2026_07_sonnet5_aa_gpqa
Result
91.11%
Confidence A
- Tested
- 2026-07-01
- Runs
- 1
Configuration
How it was measured
- Exact model
- claude-sonnet-5
- Reasoning
- max
- Agent harness
- none
- Tool access
- none (closed-book scientific QA)
- Prompting
- Artificial Analysis standard GPQA Diamond protocol
PRIMARY independent INT signal for gold sheet. Artificial Analysis GPQA Diamond under adaptive reasoning max effort.
Caveats
Known limitations
- Max effort config — higher than default chat; do not map to non-reasoning.
- GPQA is near-saturated at the frontier; use with peer comparison not absolute %.
- Single published AA point estimate (no multi-trial CI on this scrape).
Weighting
Evidence weight factors
The engine weights each evidence row by these factors (0–1) when fusing scores — higher is more trusted.
relevance0.95
quality1.00
recency1.00
health0.75
comparability0.90
reliability0.85
Benchmark
GPQA Diamond
Graduate-level science questions (biology, physics, chemistry) designed to be Google-proof; Diamond is the hardest subset where PhD experts score ~65%.
Benchmark source →