Start from what you have
Use this matrix to reach your first successful generation in one session:
- Only text → Text to Image (still) or Text to Video (motion)
- One or more photos to edit → Image to Image
- A start frame (and maybe an end frame) → Image to Video in the dashboard
- Several style or subject references, no frame roles → Multi-images to Video
- Existing clips to match motion or style → Video to Video
If your goal is a still image
Open Text to Image when you are brainstorming from scratch. Pick Nano Banana 2 for speed or GPT Image 2 for maximum fidelity.
Open Image to Image when you already have product photos, portraits, or layouts to refine. You can upload up to eight references.
- Marketing stills and social posts → Text to Image
- Background swap, style change, or composite → Image to Image
- Public tool pages: /tools/image/text-to-image and /tools/image/image-to-image
If your goal is motion
Text to Video is the fastest path when you only have a script or scene description. Choose VEO 3.1 for cinematic results or Seedance 2.0 for flexible durations (4–15 seconds).
Image to Video is best when you already composed key frames. Multi-images to Video is for mood boards with several references.
Video to Video is for matching motion from reference footage—read upload rules carefully (no real people in reference video).
Activate in under five minutes
A minimal first-run checklist:
- Sign in at /login
- Confirm credits on Dashboard or Billing
- Open one tool, write a specific 2–3 sentence prompt
- Use default duration/aspect for video on the first try
- Open Creations to download the result
Category: Getting started