Harvey's Legal Agent Benchmark

Ranks compare reported scores in the same published row; ties share a rank.

Harvey's Legal Agent Benchmark. All-pass rate across 24 practice areas; GLM-Legal improves on GLM-5.3-Flash by 12.5 percentage points.

Z.ai’s three-row comparison image, read 13 October 2026. All 15 values retain the published model identities; GLM-5.3-Flash is a comparison column, not GLM 5.3. Harvey LAB is the all-pass rate across 24 practice areas, distinct from criteria pass rate. LegalBench and PLawBench retain the percentages shown in the image; their aggregation, task subsets and judge settings are not specified. Reasoning effort, harness configuration and trial counts are not stated. These are publisher reports, kept separate from independent Vals and Artificial Analysis results. Ranks are derived only within each row, with ties sharing a rank.

z.ai/blog/glm-legal

Ranks compare reported scores in the same published row; ties share a rank.

Page 5 of Google DeepMind’s model evaluation, read 1 October 2026. Gemini runs use the API at the highest thinking setting, generally pass@1 without majority voting or parallel test-time compute; smaller benchmarks average multiple trials. Other models use their highest available published reasoning results. Google quotes Vals AI for knowledge-work and Vibe Code results, Zapier’s private-set AutomationBench leaderboard, Proximal for FrontierSWE v2, Surge AI for RiemannBench and no-tools Chartography, and the CWE-bench v1 leaderboard. Argon’s DeepSWE uses mini-swe; Terminal-Bench 4.0 is self-computed for Argon and quoted from the author board for other models. PostTrainBench v1.1 uses OpenCode with one H100 and a 10-hour budget for all models. Terminal-Bench Science uses a 6x verifier timeout for Argon. LABBench 2 uses terminal, bioinformatics tools, Python, R and internet access for all models. GraphWalks uses BFS F1 on two separate subsets: 650 items up to 128k tokens and 200 items from 256k to 1M. Agent’s Last Exam uses ALE-Claw, a 5-hour window and binary pass rate, with safety filters enabled for Argon. OSWorld 2.0 uses offline partial reward, the best of three single-attempt runs for Argon, safety filters, 1080p, 500 steps, Gemini CUA with parallel batch tools and compaction, screenshot-only observation and the 08.08 patch; Claude’s combined online/offline results are omitted. LVBench runs without tools, using 1 FPS for Gemini, 800 frames for Astra, 300 for Fable and 600 for Opus. Missing cells remain blank. These are the report’s published figures, preserved separately from this site’s independent source snapshots.

storage.googleapis.com/deepmind-media/gemini/gemini_4_argon_model_evaluation.pdf