MiMo-V2.6-Pro
XiaomiOpen WeightMIT · Usage Commercial
Description
MiMo-V2.6-Pro is Xiaomi's MIT-licensed omnimodal sparse Mixture-of-Experts model with 1.02T total parameters and 42B 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-Pro-RL. Benchmark results are Xiaomi-reported release evaluations, distinct from intermediate live-training checkpoints. UltraSpeed serves the same model at higher speed for $4.35 input and $8.70 output per million tokens.
Date de sortie
2026-09-21
Paramètres
1.0T
Longueur du contexte
1.0M
Modalités
audio, image, text, video
Radar de capacités
47
general
61
coding
100
reasoning
57
science
60
agents
80
multimodal
Classements
| Domaine | #Rang | Score | Source |
|---|---|---|---|
| Capacité agentique | 26 | 57.0 | LS |
| Classement codage | 4 | 97.0 | AA |
| Classement général | 31 | 79.0 | AA |
| Classement multimodal | 41 | 59.0 | LS |
| Science | 29 | 85.0 | AA |
Scores de benchmarks (LLM Stats)
(LLM Stats (zeroeval))Agents
Toolathlon-Verified
76.9%Aut.
SEC-bench Pro
66.3%Aut.
JobBench
62.0%Aut.
AutomationBench v1.0.6
53.1%Aut.
Agents' Last Exam
31.6%Aut.
Program Bench
26.5%Aut.
ExploitGym
17.8%Aut.
Code
CyberGym
94.0%Aut.
MiMo Cyber Bench
81.7%Aut.
MiMo Visual Coding
72.3%Aut.
DeepSWE 1.1
71.9%Aut.
MiMo Coding Bench
63.2%Aut.
ExploitBench
47.9%Aut.
General
GDPval-AA 2.1
1673.00 / 3000Aut.
AA Intelligence Index v4.3
46.3%Aut.
Multimodal
OSWorld-Verified
82.0%Aut.
Reasoning
Terminal-Bench 2.1
89.9%Aut.
Terminal-Bench 4.0
34.9%Aut.
Indices d'évaluation AA
(Artificial Analysis)Lcr(Artificial Analysis)86.3
Scicode(UIUC + Argonne National Lab (2024))60.9
Hle(Center for AI Safety + Scale AI (2025))49.4
Intelligence Index(Artificial Analysis)46.3
Scores par catégorie LLM Stats
(LLM Stats (zeroeval))Reasoning100
General100
Agents100
Multimodal80
Vision80
Safety60
Code60
Tool Calling60
Tarification
Prix d'entrée$0.435 / 1M tokens
Prix de sortie$0.87 / 1M tokens
Prix mixte (3:1)$0.544 / 1M tokens
Prix de lecture cache$0.0036 / 1M tokens
Vitesse
Tokens/sec42.9
Délai du premier token3.11s
Temps de réponse49.73s
Classement des Prix par Fournisseur
Classement des Prix par Fournisseur
14 fournisseurs
Moins cher: XiaomiPlus cher: above.dev
FournisseurEntréeSortie
1XiaomiMoins cher
$0
$0
2Kilo Gateway
$0.43
$0.87
3Requesty
$0.43
$0.87
4NanoGPT
$0.435
$0.87
5OpenRouter
$0.435
$0.87
6OpenCode Go
$0.435
$0.87
7ZenMux
$0.435
$0.87
8Vercel AI Gateway
$0.435
$0.87
9DevPass (LLM Gateway)
$0.435
$0.87
10EmpirioLabs AI
$0.435
$0.87
11LLM Gateway
$0.435
$0.87
12CrossModel
$0.47
$0.94
13AIHubMix
$0.48
$0.96
14above.dev
$0.5077
$1.0154
Comparer les prix entre différents fournisseurs API pour ce modèle.