unsloth
Qwen3 0.6B
Qwen3 0.6B is a 0.60B dense qwen3 model from unsloth. Configured context length is 40,960 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters0.60B
Context40K
Attentiongqa
Licenseapache-2.0
28 layers · 8 KV heads · head dim 128 · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| BF16 | 1.12 GB | Hugging Face file bytes |
| UD-Q8_K_XL | 0.79 GB | Hugging Face file bytes |
| Q8_0 | 0.60 GB | Hugging Face file bytes |
| UD-Q6_K_XL | 0.54 GB | Hugging Face file bytes |
| Q6_K | 0.46 GB | Hugging Face file bytes |
| UD-Q5_K_XL | 0.42 GB | Hugging Face file bytes |
| Q5_K_M | 0.41 GB | Hugging Face file bytes |
| Q5_K_S | 0.41 GB | Hugging Face file bytes |
| UD-Q4_K_XL | 0.38 GB | Hugging Face file bytes |
| Q4_K_M | 0.37 GB | Hugging Face file bytes |
| Q4_K_S | 0.36 GB | Hugging Face file bytes |
| IQ4_NL | 0.36 GB | Hugging Face file bytes |
| IQ4_XS | 0.34 GB | Hugging Face file bytes |
| Q4_1 | 0.38 GB | Hugging Face file bytes |
| Q4_0 | 0.36 GB | Hugging Face file bytes |
| UD-Q3_K_XL | 0.33 GB | Hugging Face file bytes |
| Q3_K_M | 0.32 GB | Hugging Face file bytes |
| Q3_K_S | 0.30 GB | Hugging Face file bytes |
| UD-IQ3_XXS | 0.26 GB | Hugging Face file bytes |
| UD-Q2_K_XL | 0.28 GB | Hugging Face file bytes |
| Q2_K | 0.55 GB | Hugging Face file bytes |
| UD-IQ2_XXS | 0.22 GB | Hugging Face file bytes |
| UD-IQ1_M | 0.21 GB | Hugging Face file bytes |
| UD-IQ1_S | 0.20 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Runs great | BF16 | 74–101 |
| GeForce RTX 3060 12GB | Runs great | BF16 | 93–126 |
| GeForce RTX 4060 Ti 16GB | Runs great | BF16 | 74–101 |
| GeForce RTX 4090 | Runs great | BF16 | 259–354 |
| GeForce RTX 5090 | Runs great | BF16 | 461–629 |
| Apple M4 Pro | Runs great | BF16 | 70–96 |
| Apple M5 | Runs great | BF16 | 39–54 |
| CPU only · DDR5-5600 | Runs on CPU | BF16 | 23–31 |
Smallest devices that fit it fully
- GeForce RTX 3050 6GB — Runs great, BF16, 43–59 tok/s
- GeForce RTX 2060 6GB — Runs great, BF16, 86–118 tok/s
- GeForce GTX 1660 Super — Runs great, BF16, 86–118 tok/s
- GeForce GTX 1660 Ti — Runs great, BF16, 74–101 tok/s
- Intel Arc A380 — Runs great, BF16, 48–65 tok/s
- GeForce RTX 5060 Ti 8GB — Runs great, BF16, 115–157 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (BF16).
llama.cpp
llama-server -hf unsloth/Qwen3-0.6B-GGUF:BF16 -c 8192 -ngl 28
Ollama
ollama run qwen3
LM Studio
lms get unsloth/Qwen3-0.6B-GGUF # Then load quant BF16