Qwen2.5 models
56 Qwen2.5 models on Hugging Face, from 135M to 73.4B parameters. At the precision each one is published in, the smallest needs about 0.6 GB of VRAM (turn-detector, cheapest live fit: V100) and the largest about 164 GB (Qwen2.5-VL-72B-Instruct, cheapest live fit: RTX A5000). The cheapest way to run turn-detector is $0.088/hr.
Qwen2.5 models
VRAM is for the precision the model is published in, with the same overhead and an 8,192-token context assumed on every page; see the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
