Text to video means you describe a shot in words and the model renders a moving clip from that description. On wan2-6.net you paste a prompt, choose one of the live Wan routes, review the credit cost shown next to the form, and submit. There is no editing timeline to learn: the prompt is the input, and the clip is the output.
Describe three things in one or two sentences: the subject, what it is doing, and how the camera behaves. "A woman in a red coat walks through a rainy street, slow dolly in, shallow depth of field" gives the model a subject, an action and a camera move. Prompts that only name a subject tend to produce drifting, aimless motion.
Keep the description literal. Words like "cinematic" pull some weight, but naming the light ("overcast afternoon"), the framing ("medium shot") and the pace ("slow") changes the output far more. If something unwanted keeps appearing, put it in the negative prompt instead of adding more instructions to the positive one.
Text to video works best for single, continuous shots: a subject moving through a scene, weather, product turntables, abstract motion. It is a poor fit for anything needing exact dialogue timing, on-screen text that must be legible, or two distinct shots cut together — those are better served by rendering each shot separately and editing afterwards.
Output quality varies by route and by prompt. Review what you get before you rely on it; treat the first render as a test, not a final asset.
How much does a clip cost? The credit cost is shown beside the form before you submit, based on the model, resolution and duration you chose. See pricing for what credits cost.
Can I use the result commercially? Check the terms of the model route you used; licences differ between versions.
Where do I start? Open the Wan workspace, pick the text-to-video tab, and paste a prompt. If you would rather begin from a still image, use image to video instead.