Help Center
Getting help
Everything below, plus a real human at [email protected].
Getting started
- Sign in. Use Google or Apple on the home page - new accounts include one free starter generation.
- Pick a model. The dashboard lists every enabled model with its capabilities and price - video models charge per second, image models per image (see AI Services).
- Describe your video. Write a prompt, add an image or first/last frames when the model supports them, and submit. The cost in credits is shown before you confirm.
- Wait for completion. Results stream into the grid on the page; you get an email when a video finishes or fails. Most generations complete within minutes.
- Download. Every completed video has a download button, and all of your finished videos stay available on the Videos page.
Prompting tips
- Describe one shot. A single subject, one action, one camera move reads better than a whole storyline - e.g. “slow dolly-in on a red bicycle leaning against a brick wall, morning light”.
- Name the look. Style, lighting, lens and mood words (“35 mm film grain”, “golden hour”, “soft studio light”) steer the result more reliably than adjectives about quality.
- Be explicit about motion. Say what moves and how: “leaves drift left to right”, “camera orbits the subject”. Unstated motion is chosen for you, and not always well.
- Use frames for control. On models with keyframe support, a first frame fixes composition and style; adding a last frame defines where the motion should land.
- Iterate in small steps. Change one thing per regeneration so you learn what the model responds to; keep the prompts of results you like.
Frequently asked questions
Video models are priced per second of output, and some resolutions cost more per second; image models have a flat price per image. Each model page shows the exact cost for the duration and resolution you pick before you submit, and the full table is on the pricing page. 1 credit equals $1: buy a one-time pack and the shown cost is deducted when you submit.
You are refunded automatically - the credits for that video return to your wallet exactly once, and the wallet history shows a refund row. The results grid also explains the failure reason.
Most videos complete within a few minutes. Every generation is guaranteed to reach a terminal state: completed with a download, or failed with a reason and a refund - a job can never hang forever.
Every delivered video carries a small visible logo. This is deliberate: it marks AI-generated content as such, per our AI Ethics & Safety policy, and cannot be removed.
No. Credits stay in your wallet until you spend them, and every movement (purchase, spend, refund) is listed in the wallet history.
JPEG, PNG or WebP images up to 10 MB each, used as image prompts or first/last frames on models that support them. Uploads stay private on your account.
Yes, subject to our Terms of Service and copyright policy: you must hold the rights to everything you upload, and the content must comply with the acceptable-use rules. Generated videos keep their watermark.
Paddle, our merchant of record, processes the payment and issues the receipt. Prices shown are tax-exclusive US dollars; Paddle adds your local VAT/sales tax where applicable and confirms the total at checkout. We never see or store your card details.
No. The only cookies we set are strictly necessary ones (session and CSRF protection) that keep you signed in and the forms safe. There are no analytics or advertising cookies - the privacy policy lists every cookie we set.
Email [email protected] from your account address and we will delete your account, wallet and stored files, per the privacy policy.
Still stuck?
Contact support