Qwen 3.8 (27B, IQ1_S)
Alibaba's Qwen 3.8 at 27B parameters: a dense, multimodal agentic model with native text, image, and video understanding, 262K context, flexible reasoning effort, MTP, and native tool calling. Unsloth's dynamic UD-IQ1_S is an extreme-compression build at just over 6GB, making the 27B accessible on machines with 16GB usable memory while trading more fidelity than the Q2 build for the smaller footprint.
At a glance
- Parameters
- 27B
- Quantization
- UD-IQ1_S
- Context window
- 262K tokens
- Approx. size
- 6.2 GB
- Engines
- llama.cpp
- License
- Apache-2.0 (open)
- Version
- v1.0.0 , released 2026-08-23
- Family
- qwen
Tuning
Defaults a gezel client applies out of the box, from real evaluation runs against this exact quantization.
- sampling.temperature
1- sampling.topP
0.95- sampling.topK
20- sampling.minP
0- sampling.maxTokens
12288- sampling.repetitionPenalty
1- sampling.presencePenalty
0- reasoning.thinkingBudget
4096- reasoning.enableThinking
true- reasoning.templateKwargs
{"reasoning_effort":"medium"}- profiles
- thinking-general, thinking-deep, thinking-coding, thinking-precise, instruct, creative
Model behaviors
Client-side behavior modules the catalog enables for this model (reasoning-tag handling, fabrication detection, prompt shaping).
- reasoning.strip-think-tags
- provider.merge-system-messages
- mcp.compact-tool-schemas
- provider.compact-write-transcript
- fabrication.detect-past-tense-no-tools
- turn.ollama-num-predict-bumped
- turn.preamble-folding
- turn.ramble-detection
- tools.mlx-grammar
- prompt.retrieval-first
- prompt.workspace-gestalt
- prompt.derive-by-execution
Sources
- llama.cpp (GGUF)
- unsloth/Qwen3.8-27B-GGUF ·
Qwen3.8-27B-UD-IQ1_S.gguf· UD-IQ1_S - Upstream
- https://huggingface.co/Qwen/Qwen3.8-27B
- Manifest
- View on GitHub
Tags
- alibaba
- agentic
- coding
- multimodal
- vision
- video
- tools
- reasoning
- long-context
- low-memory