Alibaba2026

Qwen3.5

Natively multimodal reasoning

Alibaba’s natively multimodal 3.5 line. Thinking and non-thinking modes for text and vision, from a sub-1B companion to a 9B daily driver. Strong default if you want one family for chat, images, and light agents.

Fits fromPhone / edge · 4 GB+
Contextup to 33K
Sizes4
RunsLocal on Delta

VisionReasoningMultilingualDaily driverText + visionCompact reasoning

Sizes

Show smaller quantizations?

Qwen3.5 0.8B

831 MB • 33k ctx • Q8_0

Qwen3.5 2B

2 GB • 33k ctx • Q8_0

Qwen3.5 4B

2.7 GB • 33k ctx • Q4_K_M

Qwen3.5 9B

5.7 GB • 33k ctx • Q4_K_M

Memory at context length

Approximate working set for Qwen3.5 0.8B as context grows.

1.2 GB

1.4 GB

16K

1.5 GB

33K

1.6 GB

Sizes

Memory at context length

Related families