GLM-5.3-FlashX-Legal

Fiche modèle · Zhipu AI · glm-5.3-flashx-legal

GLM-5.3-FlashX-Legal (GLM-Legal) is Z.ai's legal-domain variant of GLM-5.3-FlashX, the fast serving tier of GLM-5.3-Flash. It adds legal-domain mid-training and reinforcement-learning post-training across tens of thousands of environments modeled on real legal workflows (legal research, contract review, evidence analysis, document drafting), and is designed to run in agent harnesses with legal-data and Word drafting/revision MCP services. It accepts text, image, and video input with text output, has a 1M-token context window and 128K max output, and runs with always-on thinking (reasoning effort low, high, or max). Z.ai reports a 19.2% all-pass rate on Harvey's Legal Agent Benchmark (LAB), 12.5 points above the base model, independently evaluated by Vals. Announced 2026-10-12; Z.ai lists API pricing at $0.37 input, $0.075 cached input, and $1.25 output per million tokens (API access by application for enterprise customers), and the model is also listed on OpenRouter as z-ai/glm-5.3-flashx-legal.

Identité

OrganisationZhipu AI
Pays—
Famille—
Typellm
Date de sortie2026-10-12
Licenceproprietary
Licence commercialenon
Licence open sourcenon
Poids ouvertsnon
Auto-hébergement (DGX Spark)non — poids fermés
Modalitésimage|text|video
Fenêtre de contexte—
KV cache / token (Q8)—
Paramètres totaux—
Paramètres actifs (session)—
Source des paramètres actifs—
Curation des paramètres actifs—
Prix par session0.62 $ (1 000 000 tokens entrée + 200 000 sortie, prix le moins cher par fournisseur)
Tokens / tâche (Data Players)1.2 M
Tokens / tâche (Terminal-Bench)83.3 M (est.)
Ratio Terminal-Bench / Data Players69.4×
Coût / tâche (Terminal-Bench)5.21 $ (est.)
Tokens d'entraînement—
Coupure de connaissance—
Créé le2026-11-02T17:34:47.826189Z
Mis à jour le2026-11-04T06:05:15.642735Z
SourceLLM Stats
URLhttps://llm-stats.com/models/glm-5.3-flashx-legal

Prix par fournisseur

Statut du modèle (offres + auto-hébergement) : non.

Toutes sources confondues : catalogue llm-stats, routeurs EUrouter/Cortecs et catalogues directs (OVHcloud, Scaleway, Infomaniak). Les prix sont convertis en USD pour la comparaison ; le prix publié et le taux BCE utilisé sont dans l'infobulle. Les prix inconnus restent indiqués « — ».

FournisseurSourceSouverainetéEntrée (/M tok)Sortie (/M tok)QuantificationContexteLatence (s)Débit (tok/s)Statut
ZAIllm-statsnon0.37 $1.25 $————active

Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.

Souveraineté

Scores par catégorie (top_scores)

CatégorieScore
legal0.20
reasoning0.20
general0.20
agents0.20

Scores par benchmark

BenchmarkCatégorieScoreMaxNormalisé (API)VérifiéAuto-déclaréDaté du
legal-agent-benchmarklegal0.191.000.19nonoui2026-11-04T06:05:15.645581Z

Classements

CatégorieRangScoreNote conservatriceBenchmarks évaluésMéthodeDaté duProvenance
agents470.2030.941trueskill2026-11-04T08:08:47.839150Zrankings

Lecture : Note conservatrice = borne basse TrueSkill (niveau estimé − marge d’incertitude) ; Benchmarks évalués = nombre de benchmarks de la catégorie pris en compte. À performances égales, un modèle évalué sur moins de benchmarks obtient une note plus basse : la note ne se déduit pas d’un seul score. Provenance : rankings = top 50 de /rankings ; model_history = historique du modèle (/models/{id}/history), utilisé pour les rangs au-delà du top 50.

Données

Dates de curation (récupération depuis l'API) : models 2026-11-04 · benchmarks 2026-11-04 · scores 2026-11-04 · updates 2026-11-04 · account 2026-11-04 · model_history 2026-11-04 · rankings 2026-11-04