Text to video

Text to video, without the production

Text-to-video AI turns a written prompt into a moving clip. Qiroq reads your description, generates the scene and its sound, and returns a finished video you can download.

Resolutions
480p, 720p and 1080p
Clip length
5, 10 or 15 seconds
Styles
Cinematic, Documentary, Anime, 3D Render, Vlog and Product
Free credits
50 on sign-up · plans from $17/month

The prompt is your shot list. The more it reads like direction (subject, action, setting, light, lens, camera movement), the closer the result is to what you pictured.

How it works

  1. 01

    Write the prompt

    Use six parts in order: subject, action, setting, lighting, framing or lens, camera movement. Keep it under about 60 words.

  2. 02

    Draft cheaply

    Generate at a lower resolution while you tune the wording. Drafts cost a fraction of a final render.

  3. 03

    Render the winner

    Re-run the approved prompt at full resolution and download the clip.

How text-to-video works

A video model has learned how scenes, light and motion relate from large amounts of footage. It starts from noise and refines it, step by step, into frames that match your description and stay consistent from frame to frame.

You steer it with words rather than editing pixels, which is why specific, visual wording works better than feelings like 'inspiring' or 'epic'.

Prompting tips that save credits

Change one part of the prompt per attempt so you learn what each part does. Keep one subject doing one thing. Describe what the camera sees, not what the viewer should feel.

Prompts to start with

  • A young woman in a linen apron rolling out dough in a sunlit kitchen, warm morning light through a lattice window, 50mm close-up, slow dolly right
  • Runner on a mountain trail at sunrise, handheld follow shot from behind, breath visible in cold air
  • Slow ink clouds blooming in water, deep blue and gold, macro, black background

New to prompting? Read how to write prompts that produce usable video.

Good to know

  • One scene per clip, 5, 10 or 15 seconds long.
  • Complex choreography with many people and readable text are still hard for any current system.
  • Results vary between attempts; generate a few variations and keep the best.

Questions

Start now

Generate your first AI video today.

Free. No credit card. Ready in 10 minutes.

USD pricing · Cards, PayPal, Apple Pay · Tax invoices · Cancel anytime