Skip to main content
leaderboard
.cn
Toggle Sidebar
Search models or organizations…
⌘K
Explore
Frontier Models
Historical Trends
Local Models
Compare Models
Ranking Studio
Library
Models
Pricing
Organizations
Benchmarks
Data sources
Settings
Toggle Sidebar
leaderboard
.cn
Loading page content
Feedback
Loading page content
Back to models
Qwen3.5-397B-A17B
Open
Share
Add to comparison
Tool use · Reasoning
Overview
Deployment
Evaluations
Evaluation overview
Overall rank
#59
77 points · 202 models
Open-weight rank
#18
80.3 points · 120 models
Data coverage
91 items
7 sources · updated through 2026-09-02
Domain performance
Coding
#82 / 182
69.3
Research
#60 / 196
85.8
Agents
#58 / 148
72.8
Vision
#29 / 82
80.8
Long context
#36 / 95
80.8
Evaluation highlights
All evaluations
MMLU-ProX
Reasoning · Nemotron-3-Ultra Technical Report
86.4%
1
/ 22
MathVista mini
Multimodal · Qwen3.5 Technical Report
90.3%
1
/ 15
LongBench v2
Long context · Nemotron-3-Ultra Technical Report
68.9%
2
/ 27
Vals.ai Financial Agent 1.1 (without web search)
Agents · Nemotron-3-Ultra Technical Report
61.3%
1
/ 7
SWE-bench Verified
Coding · Qwen3.5 Technical Report
76.4%
23
/ 63
C-Eval
Chinese · Qwen3.6-27B Technical Report
93%
5
/ 64
Basic information
Context (input / output)
262K / 33K
Parameters
397B / 17B active
Release date
2026-02
License
Apache 2.0
Input modalities
Text · Image
Output modalities
Text
Architecture
MoE
Availability
vLLM · BF16
Deployment options
BF16 · 8 × 141 GB VRAM
Related links
Official site
HuggingFace