Generating videos with AI

You generate video clips in Zalify Canvas → Images & Videos, the same tab as AI images. Video generation is priced per second of output, in US dollars, and the rate depends on the model and resolution tier you choose — so pick both before you generate, and keep an eye on your usage balance.

Generate a video clip

1. Open Zalify Canvas and go to Images & Videos.
2. Start a new video generation.
3. Write a prompt describing the clip — subject, motion, style, and setting. The prompting advice in Generating images with AI applies to video too, with one addition: describe the motion you want ("slow pan across the product", "steam rising from the cup").
4. Choose a model, a resolution tier and a clip length. The generate button shows an estimate for that exact combination — for example "Generate video · est. $0.24" — before you commit.
5. Run the generation. Video takes noticeably longer than images — you can leave the page and come back.
6. Review the result, iterate if needed, and save keepers to your Assets library.

Models, resolution tiers and pricing

Every video model in Canvas is billed the same way: a per-second rate for the resolution tier you pick, multiplied by the clip length. Aspect ratio, prompt length and the number of reference images don't change the price. Higher tiers look better but cost more, so match the tier to where the video will be used — a small popup teaser doesn't need 1080p.

Each model page lists every tier's rate and what a typical clip costs — see the AI model pricing overview. You are only charged for renders that complete; a failed generation deducts nothing.

The usage balance is shared across everything on your plan that consumes usage — see Usage balance and auto top-up for how balances, plan allowances, and auto top-up work.

Rate limits for low-balance accounts

Accounts with a low usage balance are rate-limited on video generation. This protects you from a runaway series of expensive generations draining what's left.

If you hit the rate limit mid-project, you have two options: wait a minute, or add balance. Enabling auto top-up avoids the problem entirely — see Usage balance and auto top-up.

Keeping costs down while you iterate

Because every second of video costs money, iterate cheaply and finish expensive:

1. Nail the concept with images first. Generate still images to settle subject, style, and composition — an image costs a few cents against dollars for a clip. See Generating images with AI.
2. Draft on the cheapest model and tier. Run your prompt experiments on Seedance 2.0 at 480p until the motion and content are right.
3. Render the final at your target model and tier. Once the prompt is dialed in, generate the finished clip on the model and resolution the placement actually needs.

Saving and using your videos

Save finished clips to your Assets library so they're available across your designs — see Assets library for organizing and format details.

Troubleshooting

• Generation is blocked or queued unexpectedly: check your usage balance first — you may have crossed the low-balance threshold and been rate-limited, or the balance can't cover the estimate. Add balance or enable auto top-up in Usage balance and auto top-up.
• Clip quality is lower than expected: check which model and resolution tier the clip was generated at; drafts made at a low tier need a final render at a higher tier.
• The result ignores your brand: as with images, confirm your Brand Kit is set up.