§01·model · /models
CogVideoX 1.5
videoactiveCogVideoX License
5B text-to-video DiT by THUDM/Zhipu (CogVideoX 1.5), generating up to 1360x768. Runs via diffusers (CogVideoXTransformer3DModel + T5 + 3D causal VAE); BF16 from ~10 GB, INT8 (torchao) from ~7 GB. An image-to-video sibling (CogVideoX1.5-5B-I2V) also exists. Released under the CogVideoX License.
§02·GPUs that run this model
6 total✓ benchmarked·~ runs via recipe (not benchmarked)·— untested·✕doesn't fit