Grok 4.5
Grok 4.5 is SpaceXAI's frontier model for coding, agentic tasks, and knowledge work. Trained alongside Cursor across tens of thousands of NVIDIA GB300 GPUs, it uses large-scale reinforcement learning focused on multi-step software engineering and other technical work. The model supports text and image inputs, reasoning, function calling, structured outputs, and a 500K-token context window, and is served at 80 tokens per second.
Identité
| Organisation | xAI |
|---|---|
| Pays | — |
| Famille | — |
| Type | llm |
| Date de sortie | 2026-07-16 |
| Licence | proprietary |
| Licence commerciale | non |
| Licence open source | non |
| Poids ouverts | non |
| Auto-hébergement (DGX Spark) | non — poids fermés |
| Modalités | image|text |
| Fenêtre de contexte | 500 000 |
| KV cache / token (Q8) | — |
| Paramètres totaux | — |
| Paramètres actifs (session) | — |
| Source des paramètres actifs | — |
| Curation des paramètres actifs | — |
| Prix par session | 3.20 $ (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) | 51.5 M |
| Ratio Terminal-Bench / Data Players | 42.9× |
| Coût / tâche (Terminal-Bench) | 31.82 $ |
| Tokens d'entraînement | — |
| Coupure de connaissance | 2026-02-01 |
| Créé le | 2026-07-12T18:23:22.219967Z |
| Mis à jour le | 2026-11-04T06:05:13.420382Z |
| Source | LLM Stats |
| URL | https://llm-stats.com/models/grok-4.5 |
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 « — ».
| Fournisseur | Source | Souveraineté | Entrée (/M tok) | Sortie (/M tok) | Quantification | Contexte | Latence (s) | Débit (tok/s) | Statut |
|---|---|---|---|---|---|---|---|---|---|
| xAI | llm-stats | non | 2.00 $ | 6.00 $ | — | — | — | 80.0 | active |
Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.
SouverainetéScores par catégorie (top_scores)
| Catégorie | Score |
|---|---|
| knowledge | 42.30 |
| physics | 0.90 |
| reasoning | 13.10 |
| science | 42.30 |
| general | 9.50 |
| agents | 0.50 |
| biology | 0.90 |
| chemistry | 0.90 |
| code | 0.50 |
| tool_calling | 0.50 |
Scores par benchmark
| Benchmark | Catégorie | Score | Max | Normalisé (API) | Vérifié | Auto-déclaré | Daté du |
|---|---|---|---|---|---|---|---|
| aa-omniscience-index | knowledge | 126.00 | 200.00 | 0.63 | non | non | 2026-07-17T23:36:38.746706Z |
| artificial-analysis | general | 0.54 | 1.00 | 0.54 | non | non | 2026-07-17T23:36:38.731251Z |
| automationbench-aa | agents | 0.51 | 1.00 | 0.51 | non | non | 2026-07-17T23:36:38.741779Z |
| deepswe | code | 0.53 | 1.00 | 0.53 | non | oui | 2026-07-17T23:36:38.721283Z |
| deepswe-1.0 | code | 0.62 | 1.00 | 0.62 | non | oui | 2026-11-04T06:05:13.425920Z |
| deepswe-1.1 | code | 0.54 | 1.00 | 0.54 | non | non | 2026-11-04T06:05:13.429234Z |
| frontiercode-1.1 | reasoning | 0.42 | 1.00 | 0.42 | non | non | 2026-11-04T06:05:13.442808Z |
| gpqa | reasoning | 0.93 | 1.00 | 0.93 | non | non | 2026-07-17T23:36:38.744228Z |
| omniscience | science | 0.52 | 1.00 | 0.52 | non | non | 2026-07-17T23:36:38.749205Z |
| omniscience-non-hallucination-rate | knowledge | 0.46 | 1.00 | 0.46 | non | non | 2026-07-17T23:36:38.751687Z |
| swe-bench-pro | reasoning | 0.65 | 1.00 | 0.65 | non | oui | 2026-11-04T06:05:13.438642Z |
| swe-marathon | code | 0.29 | 1.00 | 0.29 | non | oui | 2026-11-04T06:05:13.432409Z |
| tau3-banking | agents | 0.33 | 1.00 | 0.33 | non | non | 2026-07-17T23:36:38.739380Z |
| terminal-bench-2.1 | reasoning | 0.83 | 1.00 | 0.83 | non | oui | 2026-11-04T06:05:13.435331Z |
| terminal-bench-4.0 | reasoning | 0.12 | 1.00 | 0.12 | non | non | 2026-11-04T06:05:13.445895Z |
Classements
| Catégorie | Rang | Score | Note conservatrice | Benchmarks évalués | Méthode | Daté du | Provenance |
|---|---|---|---|---|---|---|---|
| knowledge | 6 | 42.30 | 18.20 | 3 | trueskill | 2026-11-04T08:08:45.139626Z | rankings |
| agents | 44 | — | 30.29 | 9 | trueskill | 2026-10-20 | model_history |
| code | 51 | — | 31.84 | 7 | trueskill | 2026-10-20 | 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