unsloth
Apertus 8B Instruct 2509
Apertus 8B Instruct 2509 is a 8.05B dense apertus model from unsloth. Configured context length is 65,536 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters8.1B
Context64K
Attentiongqa
Licenseapache-2.0
32 layers · 8 KV heads · head dim 128 · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| BF16 | 15.0 GB | Hugging Face file bytes |
| UD-Q8_K_XL | 9.89 GB | Hugging Face file bytes |
| Q8_0 | 7.98 GB | Hugging Face file bytes |
| UD-Q6_K_XL | 6.88 GB | Hugging Face file bytes |
| Q6_K | 6.16 GB | Hugging Face file bytes |
| UD-Q5_K_XL | 5.42 GB | Hugging Face file bytes |
| Q5_K_M | 5.41 GB | Hugging Face file bytes |
| Q5_K_S | 5.23 GB | Hugging Face file bytes |
| UD-Q4_K_XL | 4.78 GB | Hugging Face file bytes |
| Q4_K_M | 4.71 GB | Hugging Face file bytes |
| Q4_K_S | 4.40 GB | Hugging Face file bytes |
| IQ4_NL | 4.37 GB | Hugging Face file bytes |
| IQ4_XS | 4.17 GB | Hugging Face file bytes |
| Q4_1 | 4.79 GB | Hugging Face file bytes |
| Q4_0 | 4.38 GB | Hugging Face file bytes |
| UD-Q3_K_XL | 4.05 GB | Hugging Face file bytes |
| Q3_K_M | 3.88 GB | Hugging Face file bytes |
| Q3_K_S | 3.43 GB | Hugging Face file bytes |
| UD-IQ3_XXS | 3.13 GB | Hugging Face file bytes |
| UD-Q2_K_XL | 3.26 GB | Hugging Face file bytes |
| Q2_K | 6.24 GB | Hugging Face file bytes |
| UD-IQ2_XXS | 2.38 GB | Hugging Face file bytes |
| UD-IQ1_M | 2.19 GB | Hugging Face file bytes |
| UD-IQ1_S | 2.08 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Tight fit | UD-Q5_K_XL | 23–31 |
| GeForce RTX 3060 12GB | Tight fit | Q8_0 | 21–28 |
| GeForce RTX 4060 Ti 16GB | Runs | UD-Q8_K_XL | 14–18 |
| GeForce RTX 4090 | Runs | BF16 | 32–44 |
| GeForce RTX 5090 | Runs great | BF16 | 57–78 |
| Apple M4 Pro | Tight fit | BF16 | 8.7–12 |
| Apple M5 | Tight fit | Q8_0 | 8.7–12 |
| CPU only · DDR5-5600 | Runs on CPU | BF16 | 2.9–3.9 |
Smallest devices that fit it fully
- GeForce RTX 3050 6GB — Tight fit, IQ4_XS, 18–25 tok/s
- GeForce RTX 2060 6GB — Tight fit, IQ4_XS, 37–50 tok/s
- GeForce GTX 1660 Super — Tight fit, IQ4_XS, 37–50 tok/s
- GeForce GTX 1660 Ti — Tight fit, IQ4_XS, 32–43 tok/s
- Intel Arc A380 — Tight fit, IQ4_XS, 20–28 tok/s
- GeForce RTX 5060 Ti 8GB — Tight fit, UD-Q5_K_XL, 36–49 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (UD-Q8_K_XL).
llama.cpp
llama-server -hf unsloth/Apertus-8B-Instruct-2509-GGUF:UD-Q8_K_XL -c 8192 -ngl 32
No matching name in the Ollama library snapshot.
LM Studio
lms get unsloth/Apertus-8B-Instruct-2509-GGUF # Then load quant UD-Q8_K_XL