CogVideoX models
3 CogVideoX models on Hugging Face, from 1.7B to 5.6B parameters. At the precision each one is published in, the smallest needs about 3.8 GB of VRAM (CogVideoX-2b, cheapest live fit: V100) and the largest about 12.6 GB (CogVideoX-5b-I2V, cheapest live fit: RTX 5060 Ti). The cheapest way to run CogVideoX-2b is $0.088/hr.
CogVideoX models
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| CogVideoX-2b | 1.7B | F16 | 3.8 GB | V100 | $0.088/hr |
| CogVideoX-5b | 5.6B | BF16 | 12.5 GB | RTX 5060 Ti | $0.110/hr |
| CogVideoX-5b-I2V | 5.6B | BF16 | 12.6 GB | RTX 5060 Ti | $0.110/hr |
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.
