unsloth
Qwen3.5 27B
Qwen3.5 27B is a 26.9B dense qwen35 model from unsloth. Configured context length is 262,144 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters26.9B
Context256K
Attentionhybrid
Licenseapache-2.0
64 layers · 4 KV heads · head dim 256 · 48 linear-attention layers · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| BF16 | 50.1 GB | Hugging Face file bytes |
| UD-Q8_K_XL | 33.1 GB | Hugging Face file bytes |
| Q8_0 | 26.6 GB | Hugging Face file bytes |
| UD-Q6_K_XL | 23.9 GB | Hugging Face file bytes |
| Q6_K | 20.9 GB | Hugging Face file bytes |
| UD-Q5_K_XL | 18.8 GB | Hugging Face file bytes |
| Q5_K_M | 18.3 GB | Hugging Face file bytes |
| Q5_K_S | 17.6 GB | Hugging Face file bytes |
| UD-Q4_K_XL | 16.4 GB | Hugging Face file bytes |
| Q4_K_M | 15.6 GB | Hugging Face file bytes |
| Q4_K_S | 14.7 GB | Hugging Face file bytes |
| IQ4_NL | 14.6 GB | Hugging Face file bytes |
| IQ4_XS | 13.9 GB | Hugging Face file bytes |
| Q4_1 | 16.0 GB | Hugging Face file bytes |
| Q4_0 | 14.6 GB | Hugging Face file bytes |
| UD-Q3_K_XL | 13.4 GB | Hugging Face file bytes |
| Q3_K_M | 12.6 GB | Hugging Face file bytes |
| Q3_K_S | 11.4 GB | Hugging Face file bytes |
| UD-IQ3_XXS | 10.7 GB | Hugging Face file bytes |
| UD-Q2_K_XL | 10.4 GB | Hugging Face file bytes |
| UD-IQ2_XXS | 7.98 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Partial offload | Q4_K_M | 4.1–5.5 |
| GeForce RTX 3060 12GB | Runs | UD-IQ2_XXS | 22–30 |
| GeForce RTX 4060 Ti 16GB | Tight fit | UD-Q3_K_XL | 11–15 |
| GeForce RTX 4090 | Tight fit | Q6_K | 24–33 |
| GeForce RTX 5090 | Tight fit | Q8_0 | 34–46 |
| Apple M4 Pro | Tight fit | UD-Q4_K_XL | 8.3–11 |
| Apple M5 | Runs | UD-IQ2_XXS | 9.2–13 |
| CPU only · DDR5-5600 | Runs on CPU | Q6_K | 2.1–2.9 |
Smallest devices that fit it fully
- GeForce RTX 3080 10GB — Tight fit, UD-IQ2_XXS, 47–64 tok/s
- Intel Arc B570 — Tight fit, UD-IQ2_XXS, 24–32 tok/s
- GeForce RTX 2080 Ti — Tight fit, UD-IQ2_XXS, 37–51 tok/s
- GeForce GTX 1080 Ti — Tight fit, UD-IQ2_XXS, 29–40 tok/s
- GeForce RTX 5070 — Runs, UD-IQ2_XXS, 41–55 tok/s
- GeForce RTX 4070 Ti — Runs, UD-IQ2_XXS, 30–41 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (UD-Q3_K_XL).
llama.cpp
llama-server -hf unsloth/Qwen3.5-27B-GGUF:UD-Q3_K_XL -c 4096 -ngl 64
Ollama
ollama run qwen3.5
LM Studio
lms get unsloth/Qwen3.5-27B-GGUF # Then load quant UD-Q3_K_XL