Qwen3.8-27B
Qwen3.8-27B is a dense 27.78-billion-parameter multimodal foundation model for coding, professional work, research, and long-horizon agents. It natively understands text, images, and video, supports configurable thinking effort and preserved reasoning across turns, and uses a 64-layer hybrid architecture with repeating Gated DeltaNet and full-attention blocks. Its native context window is 262,144 tokens and can be extended to approximately 1 million tokens. The open weights are released under Apache 2.0 and support Transformers, vLLM, SGLang, and TokenSpeed.
Identité
| Organisation | Alibaba Cloud / Qwen Team |
|---|---|
| Pays | — |
| Famille | — |
| Type | llm |
| Date de sortie | 2026-08-14 |
| Licence | apache 2.0 |
| Licence commerciale | oui |
| Licence open source | oui |
| Poids ouverts | oui |
| Auto-hébergement (DGX Spark) | oui — ≤ 107 Md (Q8/FP8) |
| Modalités | image|text|video |
| Fenêtre de contexte | 262 144 |
| KV cache / token (Q8) | 131.1 ko/token |
| Paramètres totaux | 27.8 B |
| Paramètres actifs (session) | 27.8 B |
| Source des paramètres actifs | actifs = total (modèle dense, config HF (Qwen/Qwen3.8-27B)) ; total : config Hugging Face (Qwen/Qwen3.8-27B) |
| Curation des paramètres actifs | 2026-11-04 |
| Prix par session | 1.00 $ (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 Players | 69.4× |
| Coût / tâche (Terminal-Bench) | 8.40 $ (est.) |
| Tokens d'entraînement | — |
| Coupure de connaissance | — |
| Créé le | 2026-08-14T16:00:35.294431Z |
| Mis à jour le | 2026-11-04T06:05:11.851195Z |
| Source | LLM Stats |
| URL | https://llm-stats.com/models/qwen3.8-27b |
IA Act — preuves fournisseur
| Critère | Note / 5 | Poids |
|---|---|---|
| Documentation technique | 2,5 | 5 % |
| Résumé des données d'entraînement | 2,5 | 35 % |
| Conformité droit d'auteur | 2,5 | 35 % |
| Code of Practice GPAI | 1,2 | 20 % |
| Risque systémique | 0,0 | 5 % |
Curation : 2026-10-21.
IA ActPrix par fournisseur
Statut du modèle (offres + auto-hébergement) : souverain — via AKI.IO (cortecs, eurouter), Arkane (cortecs), Berget AI AB (cortecs, eurouter), Entrim (cortecs, eurouter), IONOS Cloud (cortecs, eurouter), Infomaniak (infomaniak), OVHcloud (cortecs, eurouter, ovh), Regolo (eurouter), Scaleway (cortecs, eurouter, scaleway), Tensorix (cortecs, eurouter).
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 |
|---|---|---|---|---|---|---|---|---|---|
| AKI.IO | cortecs, eurouter | souverain | 0.17 $ | 0.73 $ | fp16 | 262 144 | — | — | — |
| Arkane | cortecs | souverain | 0.10 $ | 1.01 $ | — | 1 000 000 | — | — | — |
| Berget AI AB | cortecs, eurouter | souverain | 0.45 $ | 3.37 $ | fp8 | 262 000 / 262 144 | — | — | — |
| Entrim | cortecs, eurouter | souverain | 0.10 $ | 0.40 $ | fp8 | 262 144 | — | — | — |
| IONOS Cloud | cortecs, eurouter | souverain | 0.45 $ | 3.03 $ | int4 | 262 000 / 262 144 | — | — | — |
| Infomaniak | infomaniak | souverain | 0.49 $ | 3.67 $ | fp8 | 100 000 | — | — | — |
| OVHcloud | cortecs, eurouter, ovh | souverain | 0.45 $ | 3.03 $ | fp8 | 262 000 / 262 144 | — | — | — |
| Regolo | eurouter | souverain | — | — | — | 262 144 | — | — | — |
| Scaleway | cortecs, eurouter, scaleway | souverain | 0.67 $ | 3.70 $ | — | 256 000 / 262 000 / 262 144 | — | — | — |
| Tensorix | cortecs, eurouter | souverain | 0.40 $ | 2.42 $ | fp8 | 262 144 | — | — | — |
| DeepInfra | llm-stats | non | 0.40 $ | 3.00 $ | — | — | — | — | active |
| FriendliAI | llm-stats | non | — | — | — | — | — | — | active |
| Melious | eurouter | non | — | — | — | 262 144 | — | — | — |
Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.
SouverainetéScores par catégorie (top_scores)
| Catégorie | Score |
|---|---|
| instruction_following | 0.80 |
| math | 0.60 |
| multimodal | 0.70 |
| physics | 0.90 |
| productivity | 0.50 |
| reasoning | 0.70 |
| spatial_reasoning | 0.80 |
| structured_output | 0.90 |
| general | 0.70 |
| agents | 0.60 |
| biology | 0.90 |
| chemistry | 0.90 |
| code | 0.50 |
| tool_calling | 0.60 |
| vision | 0.70 |
Scores par benchmark
| Benchmark | Catégorie | Score | Max | Normalisé (API) | Vérifié | Auto-déclaré | Daté du |
|---|---|---|---|---|---|---|---|
| agents-last-exam | agents | 0.43 | 1.00 | 0.43 | non | oui | 2026-11-04T06:05:11.886959Z |
| androidworld | agents | 0.82 | 1.00 | 0.82 | non | oui | 2026-11-04T06:05:11.909571Z |
| babyvision | vision | 0.86 | 1.00 | 0.86 | non | oui | 2026-11-04T06:05:11.928520Z |
| charxiv-r | reasoning | 0.90 | 1.00 | 0.90 | non | oui | 2026-11-04T06:05:11.931651Z |
| claw-eval-mm | agents | 0.57 | 1.00 | 0.57 | non | oui | 2026-11-04T06:05:11.915814Z |
| coworkbench | agents | 0.71 | 1.00 | 0.71 | non | oui | 2026-11-04T06:05:11.876196Z |
| deepswe-1.1 | code | 0.42 | 1.00 | 0.42 | non | oui | 2026-11-04T06:05:11.863414Z |
| erqa | vision | 0.66 | 1.00 | 0.66 | non | oui | 2026-11-04T06:05:11.941665Z |
| gpqa | reasoning | 0.89 | 1.00 | 0.89 | non | oui | 2026-11-04T06:05:11.893154Z |
| humanity's-last-exam | reasoning | 0.31 | 1.00 | 0.31 | non | oui | 2026-11-04T06:05:11.896100Z |
| ifbench | instruction_following | 0.80 | 1.00 | 0.80 | non | oui | 2026-11-04T06:05:11.890143Z |
| job-bench | agents | 0.33 | 1.00 | 0.33 | non | oui | 2026-11-04T06:05:11.879287Z |
| livecodebench-v6 | reasoning | 0.90 | 1.00 | 0.90 | non | oui | 2026-11-04T06:05:11.899535Z |
| mathvision | math | 0.95 | 1.00 | 0.95 | non | oui | 2026-11-04T06:05:11.925488Z |
| nl2repo | code | 0.42 | 1.00 | 0.42 | non | oui | 2026-11-04T06:05:11.860504Z |
| omnidocbench-1.5 | vision | 0.91 | 1.00 | 0.91 | non | oui | 2026-11-04T06:05:11.934611Z |
| osworld-verified | multimodal | 0.84 | 1.00 | 0.84 | non | oui | 2026-11-04T06:05:11.903106Z |
| qwen-swe-bench | code | 0.79 | 1.00 | 0.79 | non | oui | 2026-11-04T06:05:11.870600Z |
| realworldqa | vision | 0.86 | 1.00 | 0.86 | non | oui | 2026-11-04T06:05:11.938260Z |
| recreationbench | agents | 0.47 | 1.00 | 0.47 | non | oui | 2026-11-04T06:05:11.912611Z |
| swe-bench-multimodal | reasoning | 0.39 | 1.00 | 0.39 | non | oui | 2026-08-14T17:15:24.688717Z |
| swe-bench-pro | reasoning | 0.62 | 1.00 | 0.62 | non | oui | 2026-11-04T06:05:11.857520Z |
| swe-mm | code | 0.39 | 1.00 | 0.39 | non | oui | 2026-11-04T06:05:11.919118Z |
| terminal-bench-2.1 | reasoning | 0.73 | 1.00 | 0.73 | non | oui | 2026-11-04T06:05:11.854342Z |
| vision2web | code | 0.63 | 1.00 | 0.63 | non | oui | 2026-11-04T06:05:11.922260Z |
| webarena-verified | agents | 0.65 | 1.00 | 0.65 | non | oui | 2026-11-04T06:05:11.906179Z |
Classements
| Catégorie | Rang | Score | Note conservatrice | Benchmarks évalués | Méthode | Daté du | Provenance |
|---|---|---|---|---|---|---|---|
| multimodal | 12 | 0.70 | 35.97 | 9 | trueskill | 2026-11-04T08:08:46.488887Z | rankings |
| agents | 45 | — | 29.46 | 12 | trueskill | 2026-10-19 | model_history |
| code | 54 | — | 30.18 | 7 | 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