Loading…
Loading…
Models / gme-Qwen2-VL
GPU memory needed (smallest version of each): gme-Qwen2-VL 2B about 6.8 GB, gme-Qwen2-VL 7B about 21.4 GB. The cheapest runs on RTX 3060 from $0.08/h right now.
2 models (2B to 7B) · ↓ 7K downloads a month · from $0.08/h on RTX 3060 right now. Each model's versions do the same job; quantized ones (AWQ, FP8, 4-bit) need less GPU memory and run on cheaper machines.
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| OriginalAlibaba-NLP/gme-Qwen2-VL-2B-Instruct | vLLM | 4.4 GB | 6.8 GB | $0.08/h RTX 3060 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| OriginalAlibaba-NLP/gme-Qwen2-VL-7B-Instruct | vLLM | 17 GB | 21.4 GB | $0.19/h RTX 3090 |
Model details come from Hugging Face and are refreshed daily; prices follow the live offers. GPU memory is GridPup's estimate: the weights, room for an 8,192-token conversation (chat models) and the runtime's own memory. Large models are split across the GPUs of one machine ("4x RTX 4090").