Grok-4 Heavy
Grok 4 Heavy is the multi-agent version of Grok 4, released alongside the standard model in summer 2025. This system spawns multiple Grok 4 agents in parallel that work independently on problems and then collaborate by comparing their solutions, similar to a study group. The agents share insights and tricks they discover, with the system intelligently combining their work rather than simply using majority voting. Grok 4 Heavy uses approximately 10x more test-time compute than regular Grok 4, enabling it to solve significantly more complex problems. On the Humanities Last Exam, it achieves over 50% accuracy on text-only problems, and it scored a perfect result on the AIME 2025 mathematics competition. The system represents a major advancement in multi-agent AI collaboration and reasoning capabilities.
Identité
| Organisation | xAI |
|---|---|
| Pays | — |
| Famille | — |
| Type | llm |
| Date de sortie | — |
| Licence | proprietary |
| Licence commerciale | non |
| Licence open source | non |
| Poids ouverts | non |
| Auto-hébergement (DGX Spark) | non — poids fermés |
| Modalités | — |
| Fenêtre de contexte | — |
| KV cache / token (Q8) | — |
| Paramètres totaux | — |
| Paramètres actifs (session) | — |
| Source des paramètres actifs | — |
| Curation des paramètres actifs | — |
| Prix par session | — |
| Tokens / tâche (Data Players) | 1.2 M |
| Tokens / tâche (Terminal-Bench) | 83.3 M (est.) |
| Ratio Terminal-Bench / Data Players | 69.4× |
| Coût / tâche (Terminal-Bench) | — |
| Tokens d'entraînement | — |
| Coupure de connaissance | 2024-12-31 |
| Créé le | 2026-05-07T16:53:39.919292Z |
| Mis à jour le | 2026-11-04T06:05:13.287124Z |
| Source | LLM Stats |
| URL | https://llm-stats.com/models/grok-4-heavy |
Prix par fournisseur
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 « — ».
Aucune donnée.
Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.
SouverainetéScores par catégorie (top_scores)
| Catégorie | Score |
|---|---|
| math | 0.80 |
| physics | 0.90 |
| reasoning | 0.80 |
| general | 0.80 |
| biology | 0.90 |
| chemistry | 0.90 |
| code | 0.80 |
| vision | 0.50 |
Scores par benchmark
| Benchmark | Catégorie | Score | Max | Normalisé (API) | Vérifié | Auto-déclaré | Daté du |
|---|---|---|---|---|---|---|---|
| aime-2025 | math | 1.00 | 1.00 | 1.00 | non | oui | 2026-11-04T06:05:13.290816Z |
| gpqa | reasoning | 0.88 | 1.00 | 0.88 | non | oui | 2026-11-04T06:05:13.294218Z |
| hmmt25 | math | 0.97 | 1.00 | 0.97 | non | oui | 2026-11-04T06:05:13.297059Z |
| humanity's-last-exam | reasoning | 0.51 | 1.00 | 0.51 | non | oui | 2026-11-04T06:05:13.300943Z |
| livecodebench | reasoning | 0.79 | 1.00 | 0.79 | non | oui | 2026-11-04T06:05:13.304514Z |
| usamo25 | math | 0.62 | 1.00 | 0.62 | non | oui | 2026-11-04T06:05:13.307735Z |
Classements
| Catégorie | Rang | Score | Note conservatrice | Benchmarks évalués | Méthode | Daté du | Provenance |
|---|---|---|---|---|---|---|---|
| math | 10 | 0.80 | 41.69 | 4 | trueskill | 2026-11-04T08:08:42.442591Z | rankings |
| code | 128 | — | 17.62 | 1 | trueskill | 2026-10-19 | model_history |
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