Qwen: Qwen3.5-122B-A10B

qwen/qwen3.5-122b-a10b

The Qwen3.5 122B-A10B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. In terms of...

Prompt $/1M

$0.260

Completion $/1M

$2.080

Context

262,144 tok

Best provider

155 tok/s · Artificial Analysis

Specs

Model id: qwen/qwen3.5-122b-a10b
Display name: Qwen: Qwen3.5-122B-A10B
Modality: text+image+video->text
Tokenizer: Qwen3
Instruct type: —
Context length: 262,144
Max completion: 65,536
Prompt $/1M: $0.260
Completion $/1M: $2.080
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
Artificial Analysis	155	999 ms	—
Alibabafastest	93	973 ms	1,408 ms
AtlasCloud	83	1,211 ms	1,571 ms
Ambient	75	1,003 ms	1,554 ms
Novita	71	1,885 ms	2,391 ms
DeepInfra	47	37,030 ms	51,998 ms
Venice	43	2,527 ms	2,927 ms
SiliconFlow	40	60,756 ms	78,770 ms