Legora's November BAR Benchmark: Muse Wonder Tops Legal AI Quality, GPT-6.2 Sol Cuts Cost 78%
shuchaobi · x · 2026-11-21
Legora released its November BAR benchmark results comparing AI models on real-world legal tasks.
Highlights:
- Muse Wonder ranks highest on overall quality and leads across four task types: analysis & answering, extraction, drafting, and editing
- GPT-6.2 Astra leads on comparison and review tasks, scoring 4% above the November model average with cost per case 20% below average
- GPT-6.2 Sol delivers above-average quality at 78% below the average cost per case
Scale AI's Alexandr Wang shared the results, calling legal work a demanding test for AI.
Related event: Meta's Muse Wonder tops Legora's November BAR legal benchmark on quality(2 posts)→
More from Models
- Sebastian Raschka: the harness is part of training, LLMs can't absorb it — rasbt · 2026-11-21
- Google's personal agent limited to one Gmail account, criticized as an own-goal — gaganghotra_ · 2026-11-21
- Jeremy Howard: tying Neon to Antigravity will limit its success — jeremyphoward · 2026-11-21
- Live Verification of Claude's Riemann Proof Hits 99%; Kontorovich Weighs In — AlexKontorovich · 2026-11-21
- GPT 6.2 Astra lands on PostTrainBench, still trailing Anthropic's Opus 5.5 — maksym_andr · 2026-11-21
- Google brings Gemini 3.5 Transcribe voice editing to the Gemini API — GeminiApp · 2026-11-21