Write the scene, pick a model, and generate a short video. If the first version feels off, adjust one part of the prompt and try again.
Ready to try?
Sign up for free and get 150 credits to start generating videos immediately.
Start generatingExample prompts
"A golden retriever running through a field of sunflowers at sunset, cinematic slow motion"
kling_5s"Aerial drone shot of a futuristic city at night, neon lights reflecting on wet streets"
kling_3_5s"A young woman painting on canvas in a sunlit studio, soft bokeh background, warm tones"
kling_10sThe basics you need to test a text-to-video idea quickly
Choose a plan that works for your creative needs
Text-to-video production notes
This page focuses on the practical side of text-to-video: how to describe a scene, when to keep prompts simple, how to choose a quality tier, and how to iterate without spending credits on random variations.
A strong text-to-video prompt tells the model what is in the frame and how it should move. Include the subject, environment, action, camera movement, lighting, mood, and visual style. A prompt like “a skincare bottle on wet stone, slow push-in camera, soft morning light, premium commercial look” gives the generator a much clearer target than “make a beauty product video.”
The first output should test whether the core idea works. Avoid asking for multiple scenes, several characters, rapid cuts, and complex story beats in one short clip. Once the baseline motion and composition look good, add one refinement at a time: a different camera move, a new lighting direction, or a clearer description of the subject.
Text prompts are best when you do not yet have a fixed visual asset. They help you discover style directions, cinematic moods, campaign concepts, and social hooks quickly. If the final output needs a precise product, logo, face, or illustration, move the winning idea into an image-to-video workflow with a reference image.
A generated clip should not only look interesting; it should communicate the idea in the first seconds. Check whether the subject is readable, motion supports the message, and the style matches where the clip will be used. This mindset makes text-to-video useful for ads, landing pages, pitches, and creator posts instead of just experimentation.
Most videos are generated within 1-3 minutes depending on the model and duration selected. Shorter clips (4-5s) are typically faster.
The current generator focuses on Gemini Omni Flash, with Seedance 2.0 available as an optional model in supported generation flows. Available models may change as provider APIs are updated.
In general, you can use generated videos for commercial projects, as long as your use complies with applicable laws, our policies, and any relevant third-party model terms.
If a generation fails, your credits are automatically refunded. You can try again with the same or different prompt.
Sign up, use the free credits, and test your first text-to-video prompt before choosing a paid plan.