Alibaba/Qwen · open-weights model
Qwen2.5-72B-Instruct
Qwen2.5-72B-Instruct is a 72.7B-parameter model by Alibaba/Qwen. Its smallest quant (Q4_K_M) needs at least 50.7GB of GPU-addressable memory at 8K context; it loads cleanly at a quality quant on 7 of the 50 machines we track.
- Parameters
- 72.7B
- Layers
- 80
- Max context
- 32K
- Attention
- GQA
- License
- apache-2.0
Config from Qwen/Qwen2.5-72B-Instruct
Quants and what they need
Measured file sizes; “needs” = weights + KV cache at 8K context + overhead.
| Quant | File size | Needs ≥ |
|---|---|---|
| Q8_0 | 77.3 GB | 80.6 GB |
| Q6_K | 64.3 GB | 67.7 GB |
| Q5_K_M | 54.5 GB | 57.8 GB |
| Q4_K_M | 47.4 GB | 50.7 GB |
What runs Qwen2.5-72B-Instruct
Every machine we track, graded at its best clean-fit quant.
Verdicts computed from published specs and measured GGUF sizes — how the math works.