Qwen3.6 Plus

Fiche modèle · Alibaba Cloud / Qwen Team · qwen3.6-plus

Qwen3.6 Plus is Alibaba's next-generation flagship model featuring a 1 million token native context window, up to 65,536 output tokens, and always-on chain-of-thought reasoning. It uses a next-generation hybrid architecture optimized for efficiency and scalability. It leads on Terminal-Bench 2.0 agentic coding (61.6), surpassing Claude 4.5 Opus, and achieves strong results on document understanding (OmniDocBench 91.2) and multimodal reasoning (MMMU 86.0). Compared to Qwen 3.5, it is significantly more decisive in reasoning, using fewer tokens on straightforward tasks with better agent stability.

Fiches modèles

Identité

OrganisationAlibaba Cloud / Qwen Team
Pays—
Famille—
Typellm
Date de sortie2026-03-31
Licenceproprietary
Licence commercialenon
Licence open sourcenon
Poids ouvertsnon
Auto-hébergement (DGX Spark)non — poids fermés
Modalitésimage|text|video
Fenêtre de contexte1 000 000
KV cache / token (Q8)—
Paramètres totaux—
Paramètres actifs (session)—
Source des paramètres actifs—
Curation des paramètres actifs—
Prix par session1.10 $ (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 Players69.4×
Coût / tâche (Terminal-Bench)9.24 $ (est.)
Tokens d'entraînement—
Coupure de connaissance—
Créé le2026-05-07T16:53:39.286621Z
Mis à jour le2026-11-04T06:05:11.160207Z
SourceLLM Stats
URLhttps://llm-stats.com/models/qwen3.6-plus
Tokens/tâche

Prix par fournisseur

Prix

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 « — ».

FournisseurSourceSouverainetéEntrée (/M tok)Sortie (/M tok)QuantificationContexteLatence (s)Débit (tok/s)Statut
Novitallm-statsnon0.50 $3.00 $————active
Together AIllm-statsnon0.50 $3.00 $————active

Souveraineté qualifiée par offre (modèle × fournisseur) ; survolez le badge pour les critères.

Souveraineté

Scores par catégorie (top_scores)

CatégorieScore
chat0.90
image_to_text0.70
instruction_following0.80
language0.90
legal0.80
long_context0.70
math0.80
multimodal0.80
physics0.80
reasoning0.80
search0.70
spatial_reasoning0.80
structured_output0.80
finance0.70
frontend_development0.80
general0.70
grounding0.80
healthcare0.80
agents0.50
biology0.90
chemistry0.80
code0.60
economics0.70
text-to-image0.80
tool_calling0.60
video0.90
vision0.80

Scores par benchmark

BenchmarkCatégorieScoreMaxNormalisé (API)VérifiéAuto-déclaréDaté du
aa-lcrreasoning0.681.000.68nonoui2026-11-04T06:05:11.246950Z
ai2dvision0.941.000.94nonoui2026-11-04T06:05:11.362878Z
aime-2026math0.951.000.95nonoui2026-11-04T06:05:11.280451Z
c-evalgeneral0.931.000.93nonoui2026-11-04T06:05:11.234499Z
cc-ocrvision0.831.000.83nonoui2026-11-04T06:05:11.359948Z
charxiv-rreasoning0.811.000.81nonoui2026-11-04T06:05:11.352558Z
claw-evalcode0.591.000.59nonoui2026-11-04T06:05:11.181361Z
countbenchvision0.98100.000.98nonoui2026-11-04T06:05:11.366683Z
deep-planningagents0.411.000.41nonoui2026-11-04T06:05:11.197776Z
dynamathmath0.881.000.88nonoui2026-11-04T06:05:11.336300Z
erqavision0.661.000.66nonoui2026-11-04T06:05:11.377235Z
finance-agent-v2agents0.411.000.41nonnon2026-11-04T06:05:11.411279Z
frontierswecode0.221.000.22nonnon2026-11-04T06:05:11.415461Z
global-piqareasoning0.901.000.90nonoui2026-11-04T06:05:11.299076Z
gpqareasoning0.901.000.90nonoui2026-11-04T06:05:11.254333Z
hmmt-2025math0.971.000.97nonoui2026-11-04T06:05:11.264343Z
hmmt-feb-26math0.881.000.88nonoui2026-11-04T06:05:11.272718Z
hmmt25math0.951.000.95nonoui2026-11-04T06:05:11.267468Z
humanity's-last-examreasoning0.291.000.29nonoui2026-11-04T06:05:11.257921Z
ifbenchinstruction_following0.741.000.74nonoui2026-11-04T06:05:11.243335Z
ifevalchat0.941.000.94nonoui2026-11-04T06:05:11.239656Z
imo-answerbenchmath0.841.000.84nonoui2026-11-04T06:05:11.276399Z
includegeneral0.851.000.85nonoui2026-11-04T06:05:11.294673Z
livebenchmath0.711.000.71nonnon2026-11-04T06:05:11.419788Z
livecodebench-v6reasoning0.871.000.87nonoui2026-11-04T06:05:11.261214Z
longbench-v2long_context0.621.000.62nonoui2026-11-04T06:05:11.251140Z
mathvisionmath0.881.000.88nonoui2026-11-04T06:05:11.329614Z
maxifegeneral0.881.000.88nonoui2026-11-04T06:05:11.312285Z
mcp-atlasreasoning0.741.000.74nonoui2026-11-04T06:05:11.207770Z
mcp-markagents0.481.000.48nonoui2026-11-04T06:05:11.204535Z
mlvuvideo0.871.000.87nonoui2026-11-04T06:05:11.394587Z
mmlongbench-docvision0.62100.000.62nonoui2026-11-04T06:05:11.356099Z
mmlu-prolanguage0.891.000.89nonoui2026-11-04T06:05:11.220601Z
mmlu-proxlanguage0.851.000.85nonoui2026-11-04T06:05:11.286993Z
mmlu-reduxlanguage0.941.000.94nonoui2026-11-04T06:05:11.223723Z
mmmlulanguage0.901.000.90nonoui2026-11-04T06:05:11.283730Z
mmmumultimodal0.861.000.86nonoui2026-11-04T06:05:11.316755Z
mmmu-provision0.791.000.79nonoui2026-11-04T06:05:11.323157Z
mmstarvision0.831.000.83nonoui2026-11-04T06:05:11.343165Z
nl2repocode0.381.000.38nonoui2026-11-04T06:05:11.187745Z
nova-63general0.581.000.58nonoui2026-11-04T06:05:11.290601Z
odinwvision0.521.000.52nonoui2026-11-04T06:05:11.373363Z
omnidocbench-1.5vision0.911.000.91nonoui2026-11-04T06:05:11.349347Z
osworld-verifiedmultimodal0.621.000.62nonoui2026-11-04T06:05:11.407027Z
polymathmath0.771.000.77nonoui2026-11-04T06:05:11.302293Z
realworldqavision0.851.000.85nonoui2026-11-04T06:05:11.339754Z
refcoco-avgvision0.94100.000.94nonoui2026-11-04T06:05:11.369894Z
screenspot-provision0.681.000.68nonoui2026-11-04T06:05:11.397879Z
simplevqavision0.67100.000.67nonoui2026-11-04T06:05:11.346279Z
skillsbenchcode0.461.000.46nonoui2026-11-04T06:05:11.184492Z
supergpqareasoning0.721.000.72nonoui2026-11-04T06:05:11.231218Z
swe-bench-multilingualreasoning0.741.000.74nonoui2026-11-04T06:05:11.167267Z
swe-bench-proreasoning0.571.000.57nonoui2026-11-04T06:05:11.170442Z
swe-bench-verifiedreasoning0.791.000.79nonoui2026-11-04T06:05:11.164014Z
tau3-benchagents0.711.000.71nonoui2026-11-04T06:05:11.190995Z
terminal-bench-2reasoning0.621.000.62nonoui2026-11-04T06:05:11.174400Z
tir-benchmultimodal0.621.000.62nonoui2026-11-04T06:05:11.403591Z
toolathlonagents0.401.000.40nonoui2026-11-04T06:05:11.201154Z
v-starvision0.971.000.97nonoui2026-11-04T06:05:11.380965Z
video-mmemultimodal0.841.000.84nonoui2026-11-04T06:05:11.387286Z
videommmumultimodal0.841.000.84nonoui2026-11-04T06:05:11.390563Z
vita-benchagents0.441.000.44nonoui2026-11-04T06:05:11.194505Z
we-mathmath0.891.000.89nonoui2026-11-04T06:05:11.333178Z
widesearchsearch0.741.000.74nonoui2026-11-04T06:05:11.217158Z
wmt24++language0.841.000.84nonoui2026-11-04T06:05:11.307744Z

Classements

CatégorieRangScoreNote conservatriceBenchmarks évaluésMéthodeDaté duProvenance
chat140.9026.721trueskill2026-11-04T08:08:43.789169Zrankings
math220.8038.9315trueskill2026-11-04T08:08:42.442591Zrankings
multimodal350.8030.8919trueskill2026-11-04T08:08:46.488887Zrankings
agents75—20.2216trueskill2026-10-19model_history
code71—27.349trueskill2026-10-19model_history

Lecture : Note conservatrice = borne basse TrueSkill (niveau estimé − marge d’incertitude) ; Benchmarks évalués = nombre de benchmarks de la catégorie pris en compte. À performances égales, un modèle évalué sur moins de benchmarks obtient une note plus basse : la note ne se déduit pas d’un seul score. Provenance : rankings = top 50 de /rankings ; model_history = historique du modèle (/models/{id}/history), utilisé pour les rangs au-delà du top 50.

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