Agentic AI · AI Reasoning · z.ai
GLM-Legal
Compiled by KHAO Editorial — aggregated from 1 source. See llms.txt for citation guidance.
◌ Single Source
Hacker News evaluated GLM-Legal across three axes: complex task execution, specialized legal reasoning, and real-world legal workflows.
Key facts
- The September 30 client report states that the project is on track and 60% complete (BR-000024), while the internal snapshot records approximately 41% completion (BR-000027)
- The September 30 “ON TRACK, 60% COMPLETE” report (BR-000024) should be read against the September 8 warning that the milestone could not honestly be met (BR-000014) and the internal ~41% snapshot
- Harvey LAB (19.2% All-Pass Rate): In real-world law firm workflows across 24 practice areas, GLM-Legal scores 19.2% and improves on its base model by 12.5 percentage points
- MSA §§2 and 5 state that payment does not waive rights and warranties survive payment (BR-000001), so continued payment alone does not resolve the dispute
Summary
Large language models are shifting from single-turn code generation toward broader professional collaboration (Professional Cowork). Legal practice demands an unbroken chain between facts, statutory doctrine, analytical reasoning, and client-ready work product. Across complex execution, legal reasoning, and real-world drafting, GLM-Legal matches frontier-tier performance while maintaining high inference efficiency:. Harvey LAB (19.2% All-Pass Rate): In real-world law firm workflows across 24 practice areas, GLM-Legal scores 19.2% and improves on its base model by 12.5 percentage points.