Qwen3.8-27B GGUF
Certified GGUF precision quantization ladder for Qwen/Qwen3.8-27B dense vision-language model.
Quantized GGUF precision variants (Q8_0, Q4_K_M, Q2_K) of Qwen/Qwen3.8-27B evaluated across standard reasoning, mathematics, and science benchmarks (ARC-MC, GSM8K, MATH-500) under reproducible seed controls.
Architecture
- Parameters
- 27.32B
- Active
- 27.32B (100%)
- Layers
- 64
- Quantization ladder
- Certified GGUF
Downloads
Exact release files
3 immutable artifacts
Qwen3.8-27B.q4_k_m.gguf
Size: 17 GB
20–24 GB VRAM (RTX 3090 / 4090 / Mac M 32GB). General conversational and daily development.
SHA-256 Checksum
ba4aeb0670e6ba5f8394524fd136062af76b6a621c21f7308059d63c49a6fadb
Qwen3.8-27B.q8_0.gguf
Size: 29 GB
32–48 GB VRAM. Production, critical reasoning, and coding.
SHA-256 Checksum
bddff381b894959327b3ad60b2b275d0e4e3ff06a187ab88f6522605d562a35b
Qwen3.8-27B.q2_k.gguf
Size: 11 GB
12–16 GB VRAM (RTX 4070 / 4080 / Mac M 16GB). Memory-constrained consumer workstations.
SHA-256 Checksum
f068c9605e58b7df9328417f8b1efccd25c5589a669bdad9d2353d7fd7180d59
Deployment
Quickstart & CLI
llama.cpp & Ollama
Stream weights directly from Hugging Face via Ollama CLI:
ollama run hf.co/Omega-Factory/Qwen3.8-27B-GGUF:Q4_K_MRun inference locally using llama.cpp with the recommended artifact:
llama-cli -hf Omega-Factory/Qwen3.8-27B-GGUF -m Qwen3.8-27B.q4_k_m.gguf -p "Explain the core mechanics of mixture-of-experts models." -n 512Launch a local OpenAI-compatible HTTP server:
llama-server -hf Omega-Factory/Qwen3.8-27B-GGUF -m Qwen3.8-27B.q4_k_m.gguf --port 8080 -ngl 99Evidence
Measured outcomes & precision ladder
Precision & Accuracy Matrix vs. FP16 Baseline
Holdout evaluation split (N=200 per category, seed 42) comparing all quant variants to unquantized baseline.
ARC-MC
94.5%
GSM8K
59.0%
MATH-500
23.0%
ARC-MC
94.5%
GSM8K
63.5%
+4.5 pp
MATH-500
22.0%
-1.0 pp
ARC-MC
95.0%
+0.5 pp
GSM8K
33.5%
-25.5 pp
MATH-500
17.5%
-5.5 pp
ARC-MC
93.0%
-1.5 pp
GSM8K
86.0%
+27.0 pp
MATH-500
10.0%
-13.0 pp
Item-Level McNemar Discordance & Flips vs. FP16 Baseline
Per-item analysis measuring individual test questions that flipped answer status relative to the unquantized FP16 baseline (N=200).
Recommended Release Rung Scoped Verification (Q4_K_M)
Formal cryptographic release marker evaluation records with 95% confidence intervals and sample counts.
ARC-Challenge Science Reasoning
qwen3.8-27b-arc-mc-v1 · accuracy_pct
Base
94.50%
Release
95.00%
Delta
+0.50 pp
Holdout evaluation set (N=200, seed 42) comparing Q4_K_M quantized release against uncompressed FP16 baseline under standard multiple-choice log-likelihood scoring.
95% CI [-3.87, 4.87] pp · 200 samples · 1 seeds · artifact ba4aeb0670e6ba5f8394524fd136062af76b6a621c21f7308059d63c49a6fadb
GSM8K Grade School Math
qwen3.8-27b-gsm8k-v1 · accuracy_pct
Base
59.00%
Release
33.50%
Delta
-25.50 pp
Holdout evaluation set (N=200, seed 42) comparing Q4_K_M quantized release against uncompressed FP16 baseline under greedy decoding. Degradation observed without fine-tuning.
95% CI [-34.95, -16.05] pp · 200 samples · 1 seeds · artifact ba4aeb0670e6ba5f8394524fd136062af76b6a621c21f7308059d63c49a6fadb
MATH-500 Advanced Mathematics
qwen3.8-27b-math500-v1 · accuracy_pct
Base
23.00%
Release
17.50%
Delta
-5.50 pp
Holdout evaluation set (N=200, seed 42) comparing Q4_K_M quantized release against uncompressed FP16 baseline under greedy decoding.
95% CI [-13.36, 2.36] pp · 200 samples · 1 seeds · artifact ba4aeb0670e6ba5f8394524fd136062af76b6a621c21f7308059d63c49a6fadb
Literature & Architecture
Foundational Research & Citations
Primary scientific publications & architecture preprints
Qwen3 Technical Report
Qwen Team, Alibaba Group
Technical report presenting Qwen3 architecture innovations, pretraining recipes, and post-training alignment for advanced reasoning and multi-turn coding.
Qwen2.5 Technical Report
Qwen Team, Alibaba Group
Foundational architecture, attention mechanisms, and pretraining scale underlying the predecessor Qwen2.5 dense models.
Which Quantization Should I Use? A Unified Evaluation of llama.cpp Quantization
Empirical GGUF Research Study
Comprehensive empirical analysis of llama.cpp GGUF quantization schemes (Q8_0, Q4_K_M, Q2_K) across reasoning benchmarks, memory footprints, and CPU/GPU throughput.
Limitations
- VISION_LANGUAGE_ARCHITECTURE — Dense Transformer with Vision-Language capabilities and native context window of 262,144 tokens.
- VRAM_ESTIMATE_NOTICE — Artifact size is disk usage, not a measured VRAM requirement. Deployment requires peak-memory profiling.
- COMPRESSED_DERIVATIVE — Quantized precision ladder of Qwen/Qwen3.8-27B under Apache 2.0 license.
Public release identity
- Release ID
- qwen3.8-27b-gguf-v1
- Release version
- 1
- Published
- 2026-10-10T04:06:20.000Z
- Model ID
- Omega-Factory/Qwen3.8-27B-GGUF
- Base model ID
- Qwen/Qwen3.8-27B
- Catalog family
- Not supplied
- Quantization rung
- Not supplied
- Supersedes
- Not supplied
- HF repository
- Omega-Factory/Qwen3.8-27B-GGUF
- HF repository URL
- https://huggingface.co/Omega-Factory/Qwen3.8-27B-GGUF
- HF revision
- 9c61b7241b8b5e68cbabfb213c262e4459c05e7d
- HF revision URL
- https://huggingface.co/Omega-Factory/Qwen3.8-27B-GGUF/tree/9c61b7241b8b5e68cbabfb213c262e4459c05e7d
- Release manifest SHA-256
- 47a3e7456d35aa717e65123ea3973c7ef61c8088ffb191cff356441ef698d249
- Catalog SHA-256
- e87595a7c9a5d0a6ae3662985e99356632737f2a673b773c938e4f93b214a99a
- Catalog generated
- 2026-10-11T12:45:00Z
- Commercial use
- ALLOWED
- License URL
- https://www.apache.org/licenses/LICENSE-2.0