Name: Qwen3.5 397B A17B
Brand: Alibaba

Qwen3.5 397B A17B is Alibaba's language model with a 262K context window and up to 66K output tokens, available from 5 providers, starting at $0.39 / 1M input and $2.34 / 1M output. Alibaba's largest Qwen3.5 MoE model with 397B total parameters and 17B activated per token, targeting maximum capability for complex reasoning and generation.

Specifications
Canonical ID	`alibaba-qwen3-5-397b-a17b`
Type	Language
Status	Active
Creator	Alibaba
Providers	Alibaba Qwen Hugging Face OpenRouter Scaleway Together AI
Context Window	262K tokens
Max Output	66K tokens
Input Modalities	ImageTextVideo
Output Modalities	Text
Reasoning Efforts	default
Parameters	397B
HuggingFace Likes	1,462
HuggingFace Downloads (30d)	710,153
HuggingFace Downloads (all-time)	2,631,436
Release Date	2026-02-16 · 5 months ago

Benchmarks
Intelligence Index	33.7 #68
Coding Index	48.2 #42
GPQA	0.9 #31
HLE	0.3 #47
IFBench	0.8 #11
Time to First Token	1.76s #432
SciCode	0.4 #77
LCR	0.7 #62
TerminalBench Hard	0.4 #37
TAU2	1.0 #20
Output TPS	58.9 #184

Capabilities

Input3/5

Text✓

Image✓

Audio·

Video✓

PDF·

Output1/5

Text✓

Image·

Audio·

Video·

Embedding·

Capabilities5/13

Reasoning✓

Adaptive Reasoning·

Function Calling✓

Parallel Function Calling✓

Structured Outputs✓

Native JSON Schema✓

Web Search·

URL Context·

Computer Use·

Code Execution·

File Search·

Prompt Caching·

Assistant Prefill·

Pricing by Provider

US Dollar ($)

Per 1M tokens

Provider	Standard		Batch
Provider	Input $ / 1M	Output $ / 1M	Input $ / 1M	Output $ / 1M
Alibaba Qwen `qwen3.5-397b-a17b`	$0.6	$3.60	$0.3	$1.80
Hugging Face `novita:qwen/qwen3.5-397b-a17b`	$0.6	$3.60	—	—
Hugging Face `scaleway:qwen3.5-397b-a17b`	$0.684	$4.10	—	—
OpenRouter `qwen/qwen3.5-397b-a17b`	$0.39	$2.34	—	—
Scaleway `scaleway/qwen/qwen3.5-397b-a17b`	$0.6	$3.60	—	—
Together AI `together_ai/Qwen/Qwen3.5-397B-A17B`	$0.6	$3.60	—	—

Cost Calculator

US Dollar ($)

Preset:

Input tokens

Output tokens

Number of calls

Cheapest Instances to Run It

Cloud GPU instances that can host Qwen3.5 397B A17B, ranked by cheapest on-demand price. The model needs about 953 GB of GPU memory at FP16 precision (estimated from its parameter count), so treat the fit as guidance rather than a guarantee.

All clouds

FP16 (full precision)

US Dollar ($)

Instance	Cloud	GPU	VRAM	Price	Cheapest region
Standard_ND96isr_MI300X_v5	Azure	8× AMD Instinct MI300X	1536 GB	$48.00/hr	westus3
p5en.48xlarge	AWS	8× H200	1128 GB	$63.30/hr	us-east-1
Standard_ND96isr_H200_v5	Azure	8× NVIDIA H200 GPU (141GB)	1128 GB	$84.80/hr	westus2
3 more instances can run Qwen3.5 397B A17B Unlock the full ranked list and FP8 / INT4 quantization with a CloudPrice subscription.

Versions

Version	Released	Context	Input / 1M	Output / 1M	Status
Qwen3.5 122B A10B	2026-02-25	262K	$0.250	$1.75	Available
Qwen3.5 35B A3B	2026-02-25	262K	$0.140	$1.00	Available
Qwen3.5 397B A17B	2026-02-16	262K	$0.390	$2.34	Current
Qwen3.7 Plus VL Instruct	—	—	—	—	Available

Model IDs

accounts/fireworks/models/qwen3p5-397b-a17b

alibaba-qwen3-5-397b-a17b

openrouter/qwen/qwen3.5-397b-a17b

qwen/qwen3.5-397b-a17b

Qwen/Qwen3.5-397B-A17B

qwen3-5-397b-a17b

qwen3-5-397b-a17b-non-reasoning

qwen3.5-397b-a17b

scaleway/qwen/qwen3.5-397b-a17b

together_ai/Qwen/Qwen3.5-397B-A17B

Qwen3.5 397B A17B

CapabilitiesAPIGET/api/v1/models/alibaba-qwen3-5-397b-a17b

Pricing by ProviderAPIGET/api/v1/models/alibaba-qwen3-5-397b-a17b/pricing

Cost CalculatorAPIGET/api/v1/models/alibaba-qwen3-5-397b-a17b/pricing/calculate?input_tokens=1000000&output_tokens=500000

Cheapest Instances to Run ItAPIGET/api/v1/models/alibaba-qwen3-5-397b-a17b/instances

VersionsAPIGET/api/v1/models?family=qwen3_5_moe

Model IDsAPIGET/api/v1/models/alibaba-qwen3-5-397b-a17b

Capabilities

Pricing by Provider

Cost Calculator

Cheapest Instances to Run It

Versions

Model IDs