Description
State-of-the-art laptop size model capable of reasoning, tools calling, and image understanding
Stats
63.1K Downloads
8 stars
Capabilities
Minimum system memory
Tags
Last updated
Updated 3 hours agobyREADME
Qwen3.8 27B
Qwen3.8 27B is a compact, deployment-friendly dense vision-language model built on the Qwen3.5 architecture. It delivers stronger performance across coding, professional work, research, and long-horizon agentic tasks.
Highlights
-
Agent Execution: stronger autonomous planning and environment-feedback handling improve reliability on complex, multi-step tasks.
-
Flexible Thinking Control: thinking is enabled by default, with
xhigh,medium, andlowreasoning-effort levels. Reasoning from previous messages is preserved by default for better continuity in agentic workflows. -
Vision-Language Understanding: native image and video understanding supports content ranging from STEM diagrams and documents to hour-scale videos.
-
Long Context: natively supports up to 262,144 tokens.
Custom Fields
Special features defined by the model author
Reasoning Effort
: select
(default=xhigh)
Controls how much reasoning the model should perform.
Enable Thinking
: boolean
(default=true)
Controls whether the model will think before replying
Preserve Thinking
: boolean
(default=true)
Preserve reasoning content in all prior assistant turns instead of only the most recent one
Parameters
Custom configuration options included with this model
Sources
The underlying model files this model uses
Based on