Apertus Models — Hardware Requirements
5 Apertus models from swiss-ai and the community, from the smallest that runs in 7.3 GB of VRAM at Q4 up to 70.6B parameters. Every row links to full quantization tables and GPU compatibility.
All Apertus Models by Size
| Model | Params | Runs from (Q4) | Context | Publisher | Quant downloads |
|---|---|---|---|---|---|
| Apertus 8B Instruct 2509 | 8.1B | 7.3 GB | 66K | ||
| Apertus 8B 2509 | 8.1B | 7.3 GB | 66K | ||
| Apertus V1.5 8B | 8.9B | 8.5 GB | — | ||
| Apertus 70B 2509 | 70.6B | 48.0 GB | 66K |
Community variants and fine-tunes(2)
| Model | Params | Runs from (Q4) | Context | Publisher | Quant downloads |
|---|---|---|---|---|---|
| Apertus 8B MeditronFO | 8.1B | 7.3 GB | 66K | ||
| Apertus 70B Instruct 2509 | 70B | 47.7 GB | 66K |
Frequently Asked Questions
- How much VRAM do I need to run a Apertus model?
- The smallest Apertus model, Apertus 8B Instruct 2509, runs from 7.3 GB of VRAM at Q4_K_M (smaller low-bit quantizations start at 2.8 GB with a noticeable quality loss). Larger family members need proportionally more — see the table above for every model.
- Which Apertus models can I run on a 16 GB GPU?
- 4 of 6 Apertus models fit in 16 GB of VRAM at Q4_K_M, including Apertus 8B Instruct 2509, Apertus V1.5 8B, Apertus 8B 2509.
- What is the most popular Apertus model to run locally?
- Apertus 8B Instruct 2509 is the most downloaded Apertus model in local-friendly quantized formats. It runs from 7.3 GB of VRAM at Q4_K_M.