Qwen: Qwen3.5-35B-A3B

qwen/qwen3.5-35b-a3b

Released Feb 25, 2026262,144 context

$0.1625/M input tokens$1.30/M output tokens

The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall performance is comparable to that of the Qwen3.5-27B.

Qwen: Qwen3.5-35B-A3B

qwen/qwen3.5-35b-a3b

Released Feb 25, 2026262,144 context

$0.1625/M input tokens$1.30/M output tokens

The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall performance is comparable to that of the Qwen3.5-27B.

Recent activity on Qwen3.5-35B-A3B

Total usage per day on OpenRouter

Prompt

69.2M

Reasoning

5.88M

Completion

927K

Prompt tokens measure input size. Reasoning tokens show internal thinking before a response. Completion tokens reflect total output length.