What is the difference between image-to-video and text-to-video?

Text-to-video generates a clip from a written prompt alone — the model invents everything. Image-to-video uses an image you provide as the first frame and animates outward — the model keeps the look you supplied. Use image-to-video whenever you need a specific product, character or style to stay accurate.

Text-to-video = full hallucination from words.
Image-to-video = animate a frame you control.
Product shots, brand shots, real people → image-to-video.
Abstract or imagined scenes → text-to-video.

When text-to-video wins

Use text-to-video when you want pure creative output: imagined worlds, fantasy creatures, stylised scenes you cannot photograph. It is also the fastest way to brainstorm composition — generate 5 options on a cheap model, pick one, then refine.

When image-to-video wins

Use image-to-video any time accuracy matters: your product, your packaging, your brand mascot, a real face. The first frame is locked to what you provide, so the model cannot warp the logo or change the colour of your bottle.

Pro workflow: generate the perfect first frame with an image model (Midjourney, Flux, Imagen) at 1080×1920. Use that as the image-to-video input. You get the best of both worlds.

FAQ

Is image-to-video more expensive than text-to-video?

Slightly — typically 20–30% more credits, because the model does extra conditioning work.