Which Legal AI model shoud I use?

Comparing objective examination scores across AI models.

Filter

Select an exam and model provider

110 questions · 110.00 points · version 13-2026-04-02
Ranking

How AI models fared on the Solicitors Qualifying Examination 1 - Functioning Legal Knowledge 1 2026

Each model is ranked by its latest published result. If scores are tied, the lower-cost model ranks higher.

4 models
#1
OpenAI: GPT-5.6 Sol
openai/gpt-5.6-sol
97.3%
107.00/110 points107 correct3 incorrect$0.3484386.1s

Source: RohasNagpal.com

#2
Claude Opus 5
anthropic/claude-opus-5
93.6%
103.00/110 points103 correct7 incorrect$0.3969486.0s

Source: RohasNagpal.com

#3
Anthropic: Claude Sonnet 5
anthropic/claude-sonnet-5
86.4%
95.00/110 points95 correct15 incorrect$0.1263334.3s

Source: RohasNagpal.com

#4
DeepSeek: DeepSeek V4 Flash
deepseek/deepseek-v4-flash
73.6%
81.00/110 points81 correct29 incorrect$0.0090494.0s

Source: RohasNagpal.com