Qwen: Qwen3.5 397B A17B

qwen/qwen3.5-397b-a17b

Qwen3.5 397B A17B is an open-weight heavyweight from Qwen — 397B parameters with 17B active per token, ranking #40 of 235 on the TryAii Score. It's strongest on agentic / tool use (#16 on Tau2-bench). API pricing starts at $0.55/$3.5 per million input/output tokens — a little under the going rate for its score class, with 14 independent hosts serving it. It's reasonably quick at roughly 37-75 tokens/sec.

Prompt $/1M

$0.550

Completion $/1M

$3.500

Context

262,144 tok

Best provider

158 tok/s · Together

Specs

Model id: qwen/qwen3.5-397b-a17b
Display name: Qwen: Qwen3.5 397B A17B
Modality: text+image+video->text
Tokenizer: Qwen3
Instruct type: —
Context length: 262,144
Max completion: 235,929
Prompt $/1M: $0.550
Completion $/1M: $3.500
Free tier: no

Capabilities

Function callingno
JSON modeno
System promptno
Visionyes
Moderatedno

Benchmark scores

Each axis is one benchmark, normalized to the best model in the field.

No benchmark data for this model.

Speed by provider

Provider	tok/s	TTFT	Total latency
Togetherfastest	158	616 ms	888 ms
StreamLakefastest	106	1,084 ms	1,477 ms
Wafer	99	1,221 ms	1,363 ms
Artificial Analysis	92	1,621 ms	—
AtlasCloud	69	1,223 ms	1,354 ms
Nebius	67	879 ms	1,145 ms
Novita	66	1,164 ms	1,233 ms
Alibaba	61	1,139 ms	1,372 ms
Venice	59	1,302 ms	1,439 ms
Phala	58	1,176 ms	1,641 ms
GMICloud	52	1,796 ms	2,128 ms
Parasail	38	472 ms	592 ms
Morph	34	810 ms	1,156 ms
DeepInfra	34	441 ms	489 ms
Chutes	25	3,823 ms	5,018 ms
DigitalOcean	15	884 ms	1,203 ms
qwenfastest	—	—	—