Modeldex
Drag to turnCard front showing

QwenTextUncommon

Qwen3.5 397B-A17B

Large open Qwen multimodal MoE for visual agents and long technical tasks

qwen3.5-397b-a17b
Type
Text
Quality
73 · Top 27% on the Epoch Capabilities Index
Context window
262,144 tokens
Max output
65,536 tokens
Input price
$0.60 / 1M tokens
Output price
$3.60 / 1M tokens
Cache read
—
Cache write
—
Parameters
403B (403,397,928,944)
Downloads
600K in the last 30 days
Likes
1.6K
Access to weights
Open download
Released
Feb 15, 2026
Knowledge cutoff
—
Takes in
Text, Image, Audio, Video
Puts out
Text
Reasoning
Yes
Tool calling
Yes
Structured output
Yes
File attachments
Yes
Access
Open weights
License
Apache 2.0
Rarity
Uncommon
Set number
784/1044
Open Qwen docsWeights on Hugging FaceQuality 73Where to run it 39Evolution 6/8Race it 102 tok/sMeet its creatureBattle Gemini 3.8 Flash

License read from the model’s Hugging Face card. Specs from models.dev, downloads from Hugging Face, last changed Oct 19, 2026. Official source: www.alibabacloud.com.

How good is it?

Public leaderboard scores for Qwen3.5 397B-A17B, refreshed daily from Epoch AI and LMArena.

Top 27% of the 277 models on the Epoch Capabilities Index: #75.

Quality is the share of a leaderboard’s models this one beats, from 0 to 100. Chat models are rated on Epoch AI’s Capabilities Index when it lists them, otherwise on LMArena; image and video models on LMArena’s image and video arenas.

Leaderboards

Benchmarks Epoch AI runs

Scores from Epoch AI and LMArena, licensed CC BY 4.0. Boards list models under their own names, so a close match is shown with the name it was scored under. Credits

Where to run it

34 providers sell Qwen3.5 397B-A17B through an API, and 5 Hugging Face hosts serve it. Prices are per million tokens, cheapest first.

API providers

ProviderInputOutputContext
NvidiaFreeqwen/qwen3.5-397b-a17bFreeFree262K
UnoRouterFreeqwen3.5-397b-a17b:freeFreeFree262K
AIHubMixCheapest paidqwen3.5-397b-a17b$0.16$0.99262K
Alibaba (China)qwen3.5-397b-a17b$0.17$1.03262K
EmpirioLabs AIqwen3-5-397b-a17b$0.17$1.03256K
Merge Gatewayqwen/qwen3.5-397b-a17b$0.17$1.03131K
OrcaRouterqwen/qwen3.5-397b-a17b$0.17$1.03262K
SiliconFlow (China)Qwen/Qwen3.5-397B-A17B$0.29$1.74262K
CrofAIqwen3.5-397b-a17b$0.35$1.75262K
Kilo Gatewayqwen/qwen3.5-397b-a17b$0.39$2.34262K
Show 24 more providers
ProviderInputOutputContext
NanoGPTqwen/qwen3.5-397b-a17b$0.39$2.34262K
SiliconFlowQwen/Qwen3.5-397B-A17B$0.39$2.34262K
TokenGoqwen/qwen3.5-397b-a17b$0.40$2.65262K
Deep InfraQwen/Qwen3.5-397B-A17B$0.45$3.00262K
NEAR AI Cloudqwen/qwen3.5-397b-a17b$0.50$3.30262K
DigitalOceanqwen3.5-397b-a17b$0.55$3.50131K
Ofoxqwen/qwen3.5-397b-a17b$0.55$3.50256K
OpenRouterqwen/qwen3.5-397b-a17b$0.50$3.60262K
AlibabaOfficialqwen3.5-397b-a17b$0.60$3.60262K
Cloudflare AI Gatewayalibaba/qwen3.5-397b-a17b$0.60$3.60262K
DevPass (LLM Gateway)qwen35-397b-a17b$0.60$3.60262K
Jalapeno CloudQwen3.5-397B-A17B$0.60$3.60262K
LLM Gatewaynovita/qwen35-397b-a17b$0.60$3.60262K
LLMTRqwen/qwen3.5-397b-a17b$0.60$3.60256K
Mixlayerqwen/qwen3.5-397b-a17b$0.60$3.60262K
Nebius Token FactoryQwen/Qwen3.5-397B-A17B$0.60$3.60262K
Novita AIqwen/qwen3.5-397b-a17b$0.60$3.60262K
Scalewayqwen3.5-397b-a17b$0.60$3.60256K
Together AIQwen/Qwen3.5-397B-A17B$0.60$3.60262K
Cortecsqwen3.5-397b-a17b$0.67$4.01262K
OVHcloud AI Endpointsqwen3.5-397b-a17b$0.71$4.25262K
Venice AIqwen3-5-397b-a17b$0.75$4.50128K
GreenPTqwen3.5-397b-a17b$0.80$4.96262K
Qiniuqwen3.5-397b-a17b——256K

Hugging Face Inference Providers

HostInputOutputContextSpeedFirst token
DeepInfraTools$0.45$3.00262K48 tok/s0.2s
NovitaTools$0.60$3.60262K57 tok/s1.4s
ScalewayTools$0.68$4.10—61 tok/s0.5s
OVHcloudFastestTools$0.71$4.25262K102 tok/s1.2s
Featherless—————

From models.dev and Hugging Face, checked daily. Resellers’ listings can lag behind their price pages, so confirm with the provider before you commit.

Face off

Current text models from other labs at a similar output price.

More from Qwen

Qwen

Qwen-Image-2.1-Turbo

SIZE7.1B
New
Image generation
TextDownloads25K/mo
ImageLikes990

Text-to-image model from Qwen, 7.1B parameters.

Open weightsQwen Research license1033/1044
Qwen

Qwen3.8 Omni Flash

CTX1M
Reasoning · Vision · Tools
TextImageAudioVideoInput$0.15/M
TextOutput$0.47/M

Qwen omni model for text, vision, audio, and multimodal agent tasks

PaidProprietary1010/1044
Qwen

Qwen-Image-2.1

SIZE7.1B
78Quality
Image generation
TextDownloads180K/mo
ImageLikes3.4K

Text-to-image model from Qwen, 7.1B parameters.

Open weightsQwen Research license1005/1044
Qwen

Qwen3.8 Flash

CTX1M
Reasoning · Vision · Tools
TextImageVideoInput$0.15/M
TextOutput$0.47/M

Qwen vision-language model for visual reasoning, documents, and agent tasks

PaidProprietary990/1044
Qwen

Qwen3.8-Flash-Next

SIZE180B
87Quality
Vision
TextImageDownloads2.5M/mo
TextLikes6.3K

Vision-language model from Qwen, 180B parameters.

Open weightsQwen Community 1.0 license985/1044
Qwen

Qwen3.8-2.4T-A95B

SIZE2.4T
Text
TextDownloads26K/mo
TextLikes1.3K

Language model from Qwen, 2.4T parameters.

Open weightsQwen3.8 Max license972/1044
Qwen

Qwen3.8-27B

SIZE28B
82Quality
Vision
TextImageDownloads7.5M/mo
TextLikes18K

Vision-language model from Qwen, 28B parameters.

Open weightsApache 2.0967/1044
Qwen

Qwen3.8 Max

CTX1M
93Quality
Reasoning · Vision · Tools
TextImageVideoPDFInput$2/M
TextOutput$6/M

2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows

PaidProprietary962/1044
Deck0/4
  1. Empty slot
  2. Empty slot
  3. Empty slot
  4. Empty slot