prices synced 2026-07-28

Qwen3.5-122B-A10B

Alibaba · Multimodal · Released Feb 2026

Compare

Qwen3.5-122B-A10B is a 122-billion-parameter multimodal model from Alibaba combining linear attention with sparse mixture-of-experts routing.

Strengths
The model handles large context windows efficiently through its hybrid linear attention and sparse MoE architecture, processing text and images with reduced computational overhead compared to dense models.
Best for
Tasks requiring substantial context windows or multimodal input processing where efficiency matters more than pure parameter density.
Limitations
Predates the Qwen3.6 series released in April 2026, which offer more recent architectural improvements; smaller than Qwen3.5 397B A17B for workloads requiring maximum capacity.

Input / 1M

$0.26

Output / 1M

$2.08

Cached input / 1M

Context window

262K

Price history

Price per 1M tokens over timeOutput $2.08; Input $0.26 as of Jun 2026.$0$0.5$1.00$1.50$2.00Output on 11 Jun 2026: $2.08Out $2.08Input on 11 Jun 2026: $0.26In $0.26Jun 2026

Snapshots

Effective Input Output Cached in Note Source
11 Jun 2026 $0.26 $2.08 Imported from OpenRouter openrouter.ai

More from Alibaba

Report a problem