MiMo-V2.6-Flash
MiMo-V2.6-Flash is Xiaomi's MIT-licensed omnimodal sparse Mixture-of-Experts model with 309B total parameters and 15B active parameters. It supports text, image, video, and audio understanding with a 1,048,576-token context window and text output. MiMo-V2.6 scales reinforcement learning across coding, general agents, visual tasks, and cybersecurity. Xiaomi's launch article is dated September 22, 2026; released weights are named MiMo-V2.6-Flash-RL. Benchmark results are Xiaomi-reported release evaluations, distinct from intermediate live-training checkpoints.
Identité
| Organisation | Xiaomi |
|---|---|
| Pays | — |
| Famille | — |
| Type | llm |
| Date de sortie | 2026-09-22 |
| Licence | mit |
| Licence commerciale | oui |
| Licence open source | oui |
| Poids ouverts | oui |
| Auto-hébergement (DGX Spark) | non — 309 Md > 107 Md |
| Modalités | audio|image|text|video |
| Fenêtre de contexte | 1 048 576 |
| KV cache / token (Q8) | 73.7 ko/token |
| Paramètres totaux | 309 B |
| Paramètres actifs (session) | 15 B |
| Source des paramètres actifs | web : https://mimo.mi.com/models/en-US/mimo-v2.6-flash |
| Curation des paramètres actifs | 2026-11-04 |
| Prix par session | 0.20 $ (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) | 83.3 M (est.) |
| Ratio Terminal-Bench / Data Players | 69.4× |
| Coût / tâche (Terminal-Bench) | 1.65 $ (est.) |
| Tokens d'entraînement | — |
| Coupure de connaissance | — |
| Créé le | 2026-09-21T21:42:06.598866Z |
| Mis à jour le | 2026-11-04T06:05:13.697538Z |
| Source | LLM Stats |
| URL | https://llm-stats.com/models/mimo-v2.6-flash |
IA Act — preuves fournisseur
| Critère | Note / 5 | Poids |
|---|---|---|
| Documentation technique | 2,5 | 5 % |
| Résumé des données d'entraînement | 5,0 | 35 % |
| Conformité droit d'auteur | 5,0 | 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) : non — hors DGX Spark (> 107 Md).
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 |
|---|---|---|---|---|---|---|---|---|---|
| Xiaomi | llm-stats | non | 0.14 $ | 0.28 $ | — | — | — | — | active |
Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.
SouverainetéScores par catégorie (top_scores)
| Catégorie | Score |
|---|---|
| knowledge | 0.60 |
| multimodal | 0.80 |
| productivity | 0.60 |
| reasoning | 0.50 |
| safety | 0.40 |
| general | 0.50 |
| agents | 0.50 |
| code | 0.50 |
| tool_calling | 0.50 |
| vision | 0.80 |
Scores par benchmark
| Benchmark | Catégorie | Score | Max | Normalisé (API) | Vérifié | Auto-déclaré | Daté du |
|---|---|---|---|---|---|---|---|
| agents-last-exam | agents | 0.28 | 1.00 | 0.28 | non | oui | 2026-11-04T06:05:13.715897Z |
| automationbench-1.0.6 | agents | 0.52 | 1.00 | 0.52 | non | oui | 2026-11-04T06:05:13.712907Z |
| cybergym | code | 0.95 | 1.00 | 0.95 | non | oui | 2026-11-04T06:05:13.734582Z |
| deepswe-1.1 | code | 0.68 | 1.00 | 0.68 | non | oui | 2026-11-04T06:05:13.700801Z |
| exploitbench | code | 0.25 | 1.00 | 0.25 | non | oui | 2026-11-04T06:05:13.740604Z |
| exploitgym | agents | 0.06 | 1.00 | 0.06 | non | oui | 2026-11-04T06:05:13.737423Z |
| jobbench | agents | 0.61 | 1.00 | 0.61 | non | oui | 2026-11-04T06:05:13.728168Z |
| mimo-coding-bench | code | 0.61 | 1.00 | 0.61 | non | oui | 2026-11-04T06:05:13.706572Z |
| mimo-cyber-bench | code | 0.77 | 1.00 | 0.77 | non | oui | 2026-11-04T06:05:13.746327Z |
| mimo-visual-coding | code | 0.71 | 1.00 | 0.71 | non | oui | 2026-11-04T06:05:13.731181Z |
| osworld-verified | multimodal | 0.81 | 1.00 | 0.81 | non | oui | 2026-11-04T06:05:13.725413Z |
| program-bench | agents | 0.26 | 1.00 | 0.26 | non | oui | 2026-11-04T06:05:13.703828Z |
| sec-bench-pro | agents | 0.47 | 1.00 | 0.47 | non | oui | 2026-11-04T06:05:13.743492Z |
| terminal-bench-2.1 | reasoning | 0.88 | 1.00 | 0.88 | non | oui | 2026-11-04T06:05:13.722373Z |
| terminal-bench-4.0 | reasoning | 0.29 | 1.00 | 0.29 | non | oui | 2026-11-04T06:05:13.719348Z |
| toolathlon-verified | agents | 0.74 | 1.00 | 0.74 | non | oui | 2026-11-04T06:05:13.709587Z |
Classements
| Catégorie | Rang | Score | Note conservatrice | Benchmarks évalués | Méthode | Daté du | Provenance |
|---|---|---|---|---|---|---|---|
| agents | 43 | 0.50 | 32.13 | 14 | trueskill | 2026-11-04T08:08:47.839150Z | rankings |
| code | 42 | 0.50 | 34.46 | 9 | trueskill | 2026-11-04T08:08:39.742311Z | rankings |
| knowledge | 35 | 0.60 | -2.66 | 1 | trueskill | 2026-11-04T08:08:45.139626Z | rankings |
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