unsloth
Qwen3.5 0.8B
Qwen3.5 0.8B is a 0.75B 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.
Parameters0.75B
Context256K
Attentionhybrid
Licenseapache-2.0
24 layers · 2 KV heads · head dim 256 · 18 linear-attention layers · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| BF16 | 1.41 GB | Hugging Face file bytes |
| UD-Q8_K_XL | 1.10 GB | Hugging Face file bytes |
| Q8_0 | 0.76 GB | Hugging Face file bytes |
| UD-Q6_K_XL | 0.72 GB | Hugging Face file bytes |
| Q6_K | 0.60 GB | Hugging Face file bytes |
| UD-Q5_K_XL | 0.56 GB | Hugging Face file bytes |
| Q5_K_M | 0.55 GB | Hugging Face file bytes |
| Q5_K_S | 0.53 GB | Hugging Face file bytes |
| UD-Q4_K_XL | 0.52 GB | Hugging Face file bytes |
| Q4_K_M | 0.50 GB | Hugging Face file bytes |
| Q4_K_S | 0.47 GB | Hugging Face file bytes |
| IQ4_NL | 0.47 GB | Hugging Face file bytes |
| IQ4_XS | 0.46 GB | Hugging Face file bytes |
| Q4_1 | 0.50 GB | Hugging Face file bytes |
| Q4_0 | 0.47 GB | Hugging Face file bytes |
| UD-Q3_K_XL | 0.46 GB | Hugging Face file bytes |
| Q3_K_M | 0.44 GB | Hugging Face file bytes |
| Q3_K_S | 0.41 GB | Hugging Face file bytes |
| UD-IQ3_XXS | 0.37 GB | Hugging Face file bytes |
| UD-Q2_K_XL | 0.39 GB | Hugging Face file bytes |
| UD-IQ2_XXS | 0.31 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Runs great | BF16 | 98–134 |
| GeForce RTX 3060 12GB | Runs great | BF16 | 122–167 |
| GeForce RTX 4060 Ti 16GB | Runs great | BF16 | 98–134 |
| GeForce RTX 4090 | Runs great | BF16 | 343–467 |
| GeForce RTX 5090 | Runs great | BF16 | 609–831 |
| Apple M4 Pro | Runs great | BF16 | 93–127 |
| Apple M5 | Runs great | BF16 | 52–71 |
| CPU only · DDR5-5600 | Runs on CPU | BF16 | 30–42 |
Smallest devices that fit it fully
- GeForce RTX 3050 6GB — Runs great, BF16, 57–78 tok/s
- GeForce RTX 2060 6GB — Runs great, BF16, 114–156 tok/s
- GeForce GTX 1660 Super — Runs great, BF16, 114–156 tok/s
- GeForce GTX 1660 Ti — Runs great, BF16, 98–134 tok/s
- Intel Arc A380 — Runs great, BF16, 63–86 tok/s
- GeForce RTX 5060 Ti 8GB — Runs great, BF16, 152–208 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (BF16).
llama.cpp
llama-server -hf unsloth/Qwen3.5-0.8B-GGUF:BF16 -c 8192 -ngl 24
Ollama
ollama run qwen3.5
LM Studio
lms get unsloth/Qwen3.5-0.8B-GGUF # Then load quant BF16