Fable 5 Tops Legal Eval, GLM-5.2 Offers Best Value
ArtificialAnlys · x · 2026-07-08
Artificial Analysis's Harvey legal evaluation shows that topping the leaderboard is expensive. Claude Fable 5 leads with a 14.2% full pass rate but is also the most expensive model at $18.9 per task, higher than Sonnet 5's $11.8; Opus 4.8 costs $8.2 per task. In terms of cost-effectiveness, DeepSeek V4 Flash (max) passed 1.7% of tasks for only $0.08, while GLM-5.2 (max) matched Opus 4.8 at roughly 15% of the cost ($1.3 per task).
Related event: Artificial Analysis Releases Harvey Legal Agent Benchmark Results(8 posts)→
More from Models
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- GPT-5.6 writes well but is instantly forgettable, user complains — BasedRaddka · 2026-09-11