Dirk-Qwen3.8-27B-GGUF
With our Sharp chat template, MTP, and vision baked in, the model answers lean and stays on-task out of the box. No template wrangling: download, point llama.cpp at it, go. If you want it to think deeper, set the effort level through chattemplatekwargs:
Params
27.0 B
Context
—
Downloads 30d
1.3 M
Likes
246
Download history
daily snapshots · 31 days
▲ 434 K in the last 30 days (50.4%)
1.5 M912 K
Sep 25Oct 5Oct 15Oct 24
1.5 M862 K
Sep 24Oct 4Oct 14Oct 24
Can you run it?
Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.
| File | Quant | Size | Est. VRAM | Verdict on RTX 4090 · 24 GB |
|---|---|---|---|---|
| Dirk-Qwen3.8-27B-GSQ-RCO-IQ2_XS.gguf | IQ2_XS | 8.8 GB | 14.2 GB | ✅ Runs comfortably |
| Dirk-Qwen3.8-27B-GSQ-RCO-IQ2_S.gguf | IQ2_S | 9.6 GB | 15.1 GB | ✅ Runs comfortably |
| Dirk-Qwen3.8-27B-UD-Q2_K_XL.gguf | Q2_K_XL | 9.8 GB | 15.4 GB | ✅ Runs comfortably |
| Dirk-Qwen3.8-27B-GSQ-RCO-IQ3_XXS.gguf | IQ3_XXS | 10.4 GB | 16.0 GB | ✅ Runs comfortably |
| Dirk-Qwen3.8-27B-GSQ-RCO-IQ3_S.gguf | IQ3_S | 12.1 GB | 17.9 GB | ✅ Runs comfortably |
| Dirk-Qwen3.8-27B-UD-Q3_K_XL.gguf | Q3_K_XL | 13.1 GB | 19.0 GB | ✅ Runs comfortably |
| Dirk-Qwen3.8-27B-UD-IQ4_XS.gguf | IQ4_XS | 14.3 GB | 20.2 GB | ⚠️ Tight — reduce context |
| Dirk-Qwen3.8-27B-UD-Q4_K_M.gguf | Q4_K_M | 16.5 GB | 22.7 GB | ⚠️ Tight — reduce context |
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.
Specifications
- Parameters
- 27.0 B
- Licence
- apache-2.0
- First seen on the Hub
- 2026-08-14
- Base model
- Qwen3.8-27B
- Training datasets
- undisclosed
- Added to our catalog
- 2026-09-24
Compare with any image-text-to-text model