Inference
Artificial Analysis

Artificial Analysis Harvey LAB-AA

An independently evaluated legal-agent benchmark from Artificial Analysis.

ModelsModels with a result.12
Top resultBest result on this test.94.6%Kimi K3
Top-3 spreadPoints from first to third.1.1 pts
YearYear of release.2026

Result and price

020406080100$1$3$10$30$100RESULTOUTPUT PRICE PER 1M TOKENS · LOG SCALE

Results

12 results
#ModelResultThe score from the source.
01Kimi K394.6%
02Claude Fable 593.6%
03Claude Opus 593.5%
04Claude Fable 5.193.0%
05Claude Opus 5.591.2%
06Gemini 3.7 Flash90.7%
07MiniMax M388.4%
08GPT-5.6 Luna87.9%
09GPT-5.6 Sol87.2%
10Nemotron 3 Ultra81.7%
11Mistral Medium 3.5 128B69.1%
12Grok 4.719.6%