A still photograph gaining motion in an image-to-video workflow
Image to Video

Grok Imagine Image to Video: Animate Any Still

Grok Imagine image to video animates a still photo into a 4-15 second clip in up to 2K. Upload a starting frame, optionally add an ending frame for transitions, then describe the motion. It is the most controllable workflow in the studio — the model keeps your composition and adds movement.

Updated 2026-10-13Jump to FAQ

First frame, last frame, and what each one controls

The first frame locks composition, subject, and lighting. The last frame is optional and defines where the shot lands — ideal for transitions, transformations, and loops.

  • Single still: the whole clip grows from one image. Best for portraits, product shots, and scenes where you want subtle, believable movement.
  • First + last frame: the model bridges two images. Use it for before/after reveals, day-to-night transitions, and match cuts.
  • No last frame: the model extrapolates the motion freely — more creative freedom, less destination control.

Image requirements at a glance

ItemRequirement
FormatsJPG, JPEG, PNG, WEBP, HEIC, or HEIF
File sizeUp to 30 MB per image
Dimensions256-5760 px, aspect ratio 0.4 to 2.5
Reference imagesUp to 9 per generation
Total request sizeUp to 64 MB across all materials

Animate a photo in four steps

  1. Upload your imageDrag in a photo or click to upload. The clearer the source, the cleaner the motion — sharp subjects on simple backgrounds animate best.
  2. Describe the motion, not the pictureThe image already defines the scene. Your prompt should only cover movement: what shifts, how the camera behaves, and the pace.
  3. Set duration, resolution, and shapeA 5-second 768P pass is a cheap way to check the motion. Re-run the keeper at 2K, and pick an aspect ratio that matches your source.
  4. Generate and compare variantsChange one variable per run — camera move, pace, or a new last frame — so you always know what caused the difference.

How to write motion prompts for a still

Think in verbs and camera terms: drifting, rising, sweeping, pushing in, pulling back, orbiting, hand-held sway. Name the pace — slow, gentle, gradual — because the model matches the energy of the motion to your adverbs.

  • Portrait: “A slow push-in; hair moves gently in a light breeze; a soft blink.”
  • Product: “The camera orbits 20 degrees; a highlight sweeps across the surface; no background movement.”
  • Landscape: “Clouds drift left to right; the light shifts warmer; the camera stays locked off.”
  • Poster or artwork: “Subtle parallax; the foreground separates from the background; embers float upward.”
Do not re-describe the image: If the prompt repeats what the photo already shows, the model spends capacity on details you already have. Motion is what you are buying — spend the words there.

Common image-to-video problems and what to change

SymptomLikely causeFix
Motion looks like a slow zoom on the whole frameNo explicit motion describedAdd subject-level motion plus a camera move, for example “hands move, camera slowly drifts right”
Subject melts or deformsBusy source or subject too smallUse a cleaner frame, keep the subject large, and shorten the move
The result ignores my last frameThe two frames are too differentBring the frames closer in framing and lighting, or bridge them with two generations
The clip feels staticAdverbs are too weakAsk for motion in measurable terms: “walks three steps”, “turns halfway”

Frequently asked questions

What is Grok Imagine image to video?+

It is a workflow that animates a still image into a short video. You upload a required first frame, optionally add a last frame, describe the motion, and generate a 4-15 second clip in 768P or 2K.

Can I use a last frame without a first frame?+

No — the first frame is required, and the last frame is optional. If you only have one image, use it as the first frame; the model will extrapolate the motion from there.

What image formats can I upload?+

JPG, JPEG, PNG, WEBP, HEIC, or HEIF, up to 30 MB each, with dimensions from 256 to 5760 px. Up to 9 reference images can be combined in one generation.

How do I keep a character consistent across clips?+

Anchor every generation to the same first frame and change only the motion description. For stronger consistency, reuse the character image as a reference in a multimodal generation.

Why does the animated clip look slightly different from my photo?+

Generation re-renders the frame, so small lighting and color drifts are normal. Gentle motion, shorter clips, and a cleaner source image keep the result closest to the original.

Does image to video cost more credits than text to video?+

Credit cost follows the duration and resolution you choose, not the workflow. The generate button always shows the estimate before a run.

Related guides

Text to Video

Grok Imagine Text to Video: Prompts That Work

How Grok Imagine text to video works, the six-part prompt formula that keeps motion coherent, and the mistakes that ruin otherwise good clips.

Read guide Examples Gallery

Grok Imagine Video Examples: 6 Prompts to Copy

Six Grok Imagine video examples with the prompt structure behind each — product ads, cinematic shots, posters, character motion, food, and vertical video.

Read guide Moderation Guide

Grok Imagine Moderated: Meaning and Fixes

Why Grok Imagine videos get flagged as moderated, the prompts and references that trigger it, and practical rewrites that get your generation approved.

Read guide Limits and Quotas

Grok Imagine Video Limits: Length, 2K & Resets

Every Grok Imagine video limit in one place — 4-15 second clips, 768P vs 2K, 9 image and 12 material caps, and what happens when you hit a limit.

Read guide

Bring Your Stills to Life

Upload a photo, describe the motion, and export a 4-15 second clip in up to 2K.

See Plans
Grok Imagine Image to Video: Animate a Still