Image-to-video models trending today
1 of the 40 models on Hugging Face’s trending list carry this task tag today, in Hugging Face’s order, with entries held back when likes are not matched by downloads.
Collected · snapshot shipped with the site build, not a live feed
1 of 1 models · Hugging Face order
Number 1: Lightricks/LTX-2.5
Image to video
No. 5 by Hugging Face trending score (408) · +13 likes since 2026-10-23 · 2,026,951 downloads in the last 30 days.
2.03M downloads in 30 days · 9,066 likes
Other (custom terms)· read the terms
Positions follow Hugging Face’s trending order with 0 of 40 entries held back for likes that downloads do not confirm. Sorting by another column changes the order on your screen only.
Source: Hugging Face public API · updated
About image-to-video models
Image-to-video models take a still picture, usually with a prompt, and generate a short clip that starts from it. They are among the largest open models by compute needed, and the trending list tends to hold one or two base models plus adapters for specific camera moves or effects.
Before you pick one
Check memory before anything else
Video models hold many frames in memory at once. The model card states the graphics memory needed at each resolution and clip length; quantised copies lower it at some cost in quality.
Licences are usually custom
Open video models are commonly released under a company licence tagged "other", sometimes with a revenue threshold for commercial use.
Clip length is a model property
Each model generates up to a fixed number of frames. Longer videos are made by chaining clips in a workflow, which is a tool feature, not something the weights do alone.
Questions readers ask
Can I run an image-to-video model on a laptop?
Generally not at full precision. Quantised checkpoints and low resolutions make some of them run on a single consumer graphics card; the model card gives the tested configurations.
Why are there so few entries in this task?
Few organisations can train a video model, so the list holds a handful of base models. The rest of the activity is in adapters, which may be tagged with other video tasks.
Is audio generated too?
Only when the model card says so. Some recent video models generate sound with the picture; most do not.