Qwen 3.8 across three opponents
98%177 of 180 equal or better
The percentage combines 15 wins and 162 functional ties within this series.
Evidence ยท Series 2
Series 2 source collection
60 legal matters, Q21 through Q80, with prompts, keys, four model outputs, six pairwise judge records per matter, code, and supporting documentation.
Outcome distribution
Qwen 3.8 judgments against all three opponents.
A/B assignment authority: Stored assignments in the verdict records are authoritative. The bundled deterministic judge helper is historical reference tooling. Read the disclosure.
Benchmark materials
- Matter prompts60 files
- Sealed keys60 files
- Model outputs240 answers
- Blind judge records360 pairings
- Original three-model generation records60 matter files
- Qwen 3.8 generation records80 matter files
Code and documentation
- Local runner reference codePython
- Hosted runner reference codePython
- Historical/reference judge helperStored assignments are authoritative
- Battery splitter reference codePython
- Exact judge prompt usedByte-identical across both series
- Current evidence hash recordSHA-256