Models

← All models

Qwen3.5-Flash

Alibaba · 2026-02-24

Qwen3.5-Flash is Alibaba's fast, cost-efficient hosted production model aligned with Qwen3.5-35B-A3B, a sparse MoE model with hybrid Gated DeltaNet and gated-attention layers; Qwen describes the corresponding open model as having 35B total and 3B activated parameters. Released on February 24, 2026, it accepts text, images, and video, returns text, provides a 1,000,000-token context window, and supports switchable thinking and non-thinking modes.

Benchmark scores