Z.ai GLM-Legal — legal-domain agentic model
Z.ai (Zhipu) launched GLM-Legal, a legal-domain model built on GLM-5.3-Flash (320B MoE, 18B active, 1M context) with legal-domain mid-training and reinforcement learning for contract review, evidence analysis, and compliance research; Z.ai reports a 19.2% all-pass rate on Harvey's Legal Agent Benchmark, roughly triple its base model.
Why it matters
A frontier lab shipping a standalone vertical model rather than an add-in: 12.5-point LAB gain over the base model, gated beta for law firms, planned AutoClaw integration — domain post-training as a product strategy.
What you could build with it
A legal-tech startup serving mid-size firms could build a contract-triage service on GLM-Legal's API, routing routine NDAs and vendor agreements through the domain-tuned model while escalating edge cases to human lawyers, cutting review turnaround from days to hours.
Does it hold up?
Early: announced Oct 12, Product Hunt-listed Oct 15, access by application only with no published pricing and no open weights; the 19.2% LAB score is vendor-reported with independent Vals AI leaderboard confirmation at 19.17%.
Built with Z.ai GLM-Legal
- Z.ai Debuts GLM-Legal, a 320B Legal Agent Built on GLM-5.3-Flash With 19.2% Harvey LAB Scorearticle · Launch coverage: 320B/18B-active MoE, 1M context, gated design-partner beta, AutoClaw desktop integration planned.
- GLM-Legal Opens Gated Beta With 19.2% LAB Claimarticle · Skeptical take: the LAB claim warrants investigation, HN discussion questions benchmark-vs-practice gaps, and access stays application-gated with no open weights.
Learn more
First spotted on article: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74StepFun Step-5-Preview — open BF16 weights releasedStepFun published the full BF16 weights for its flagship Step-5-Preview agentic model (600B-parameter sparse MoE, 27B…models · JEV 0.73ElevenLabs Eleven v4 + v4 Turbo: new emotive TTS architecture with a 100ms real-time variantElevenLabs released Eleven v4 and Eleven v4 Turbo on September 28: a new text-to-speech architecture with inline…models · JEV 0.72Mistral Large 4 (Le Chonk): 1.05T-param open-weight MoEMistral's new flagship: a 1.05T-parameter mixture-of-experts multimodal model with 49B active params and 1M-token…models · JEV 0.72VoiceStudio: open-source, fully-local ElevenLabs alternative with voice cloning, dubbing and dictation in 646 languagesVoiceStudio is an open-source, fully-local voice AI studio: voice cloning from a 3-second sample, voice design, video…models · JEV 0.71JetBrains Mellum 2.1: 12B MoE coding model trained for agentsJetBrains released Mellum 2.1, a 12B MoE model (2.5B active parameters, 128K context) under Apache 2.0, trained mainly…models · JEV 0.69
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.