How do I start an AI video project?

Start with the desired subject, action, setting, camera treatment, lighting and sound. Then use the workspace to check the models and options that are genuinely ready for your account before requesting a price.

Can I use an image or other references?

The workspace shows the input modes and reference controls supported by the model available to your account. Use only media you own or are authorized to use.

A generation is priced by the server before submission. Review the exact maximum price and confirm it before the request is sent.

AI Creator / Video

Create an AI video from a prompt.

Describe the moment, add references when the selected model supports them, and keep the creative decisions—length, frame, sound and price—in one place.

Prompt-led planningOwned-reference workflowExact quote before submitAccount-specific availability

AI Video GeneratorThe live ClipSpeedAI workspace

Open full workspace ↗

Loading your creator workspace

The tool above is the real workspace. Sign in there to see models, availability and the server price for your account. Open it in a full window.

AI video workflow

Make the prompt do the directing.

A useful video prompt is not a pile of adjectives. It gives the generator a job: who or what the viewer sees, what changes over time, how the camera observes it, and what must stay consistent. That makes it easier to judge whether a first render misses the idea—and what to revise next.

Write the visible action

Name the subject, environment and one clear progression. “A runner in rain” is a scene; “the runner looks back, accelerates and disappears into the tunnel” is a beat.

Choose the frame on purpose

Pick a frame for its destination. A vertical composition gives room for a person, product or action in a short-form feed; a wide composition gives setting and movement more room to read.

Review before you render

Check the selected model, duration, output choices and exact server quote. A concise, intentional request is easier to evaluate than a vague prompt with every setting changed at once.

What to include in an AI video prompt

Start with the shot’s non-negotiables: subject, setting, action, framing and light. Add a camera move only when it changes the experience—such as a slow push toward a reaction, a locked-off product reveal or a handheld creator point of view.

Our AI video prompt guide breaks that structure into practical examples, while the camera-movement guide helps turn camera language into something specific enough to test.

Concept footage

Turn a short creative direction into a testable motion concept before a larger campaign commits to the idea.

Product moments

Use an owned product reference to describe the setting, movement and attention beat around a product instead of relying on generic stock language.

Hook shots

Plan a striking first visual that can introduce a short-form post, then edit it with your original footage and message.

Music-led scenes

Specify motion, pacing and atmosphere together when a visual needs to land alongside a song or a spoken track.

Questions creators ask

AI video generator FAQ

Which model should I choose?

Choose only from the models the workspace marks ready for your account. Availability is deliberately shown inside the product because a display name alone is not a promise that a model can be quoted and delivered today.

How long should an AI video prompt be?

Long enough to remove important ambiguity, not so long that it contradicts itself. Establish the subject, action, camera, setting and a few visual constraints; then run one controlled change at a time.

Can I use generated video commercially?

Commercial use depends on your prompt, inputs, the provider terms and the place where you publish. Keep rights to your source media, avoid impersonation, and review the commercial-use guide before you publish.