AllVideoAIAI video help center
Get direct answers about AI video models, text and image inputs, references, creator credits, generation time, privacy, downloads, commercial use and content safety.
A multi-model AI video generator is a private workspace for creating videos from text prompts, images, source videos, audio and creative references. Available inputs and controls depend on the selected model.
Supported workflows include text-to-video, image-to-video, reference-to-video and video editing when the selected model provides those capabilities.
The model picker lists currently published model families and variants. Only models with a compatible, server-validated input schema are available for generation.
Generation settings change because each AI video model supports a different combination of duration, aspect ratio, resolution, seed, negative prompt and reference inputs.
Reference-to-video is an AI video workflow that uses approved images, videos or audio to guide subject, style, motion or continuity. Exact behavior depends on the model.
Yes. When the selected model supports them, you can upload images, source video or audio through private, account-aware storage flows.
Yes. Where supported, you can request 1, 2 or 4 outputs. Each result is tracked as an independent job, so one failure does not stop the entire batch.
AI video credits are quoted before generation. Credits are frozen when a job starts, settled for successful outputs and released for failed outputs.
No. The server quote shown at submission is the amount charged for that generation; later provider-side cost changes are not added to the job.
AI video generation time varies by model, duration, resolution, queue load and output count. Progress is saved so you can leave and return later.
Yes. Batches, individual jobs, provider attempts and completed outputs are stored with your workspace, so refreshing or signing in again does not erase progress.
Yes. Generation inputs and completed videos use private, account-aware storage. Temporary processing links may be used during generation as described in the Privacy Policy.
A failed child job releases its frozen credits independently. Other jobs in the batch can still finish, and the batch clearly shows partial success.
Possibly. Commercial-use rights depend on the model, provider terms, your inputs, the output and the intended use. Seek legal advice for regulated or high-value campaigns.
Use only material you own or are authorized to use. You are responsible for copyright, trademarks, music, voices, likenesses, privacy and publicity rights.
Prompts, inputs, outputs and account activity may be screened for prohibited sexual content, exploitation, violence, fraud, infringement and other unsafe uses.
Completed videos are copied to private account storage before they are marked complete. You can download them through an authorized link in generation history.
A preview or demo workflow may be available without third-party generation. Live video creation requires an available model and enough creator credits.