← Claude Opus 4.8 · API · High effort (default)
Evidenceacceptedindependent_lab
AA-Omniscience
Artificial Analysis — AA-Omniscience (Opus 4.8 max)
ev_2026_07_opus48_aa_omniscience
Result
27.43
Confidence A
- Tested
- 2026-06-01
- Runs
- 1
Configuration
How it was measured
- Exact model
- claude-opus-4-8
- Reasoning
- max
- Agent harness
- none
- Tool access
- AA default
- Prompting
- AA Omniscience protocol
PRIMARY WIS — AA-Omniscience calibration.
Caveats
Known limitations
- Index is not percent accuracy; show acc + hallucination rates.
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
recency0.95
health0.85
comparability0.90
reliability0.85
Benchmark
AA-Omniscience
Knowledge reliability benchmark that jointly scores accuracy and hallucination / overconfidence. Rewards calibrated abstention rather than pure guess accuracy.
Benchmark source →