en Image to Video is arguably Runway's most practical and controllable generation mode. By starting with a specific still image — whether a Midjourney generation, a photograph, or an illustration — you have precise control over the starting frame. This eliminates the unpredictability of pure text-to-video and makes it the preferred mode for professional workflows where consistency matters.
Start with a high-quality still image that captures the scene you want to animate. For best results, use images with clear subjects, good lighting, and a composition that allows for natural motion.
Upload your image to Runway's Image to Video mode. Write a motion description focusing on what should move and how: "the candle flame flickers gently", "the model's hair blows in the wind", "slow camera push forward".
For complex scenes, use the Motion Brush to paint specific areas and define their motion direction. This gives you surgical control over which elements move and which stay still.
Generate the animated clip and review the motion quality. If specific elements aren't moving as intended, use Motion Brush to correct them. Generate variations until you achieve the desired result.
Brand animating product shots for social media
Upload a perfume bottle product shot. Motion: "the bottle rotates slowly clockwise, subtle light reflections moving across the glass surface, gentle ambient particles floating in the background"
Real estate developer animating architectural renders
Upload an exterior architectural render. Motion: "slow aerial crane shot rising above the building, trees swaying gently in the breeze, clouds moving slowly across the sky"
Artist animating a digital portrait
Upload a portrait illustration. Motion: "the subject's hair moves gently in a breeze, subtle breathing motion in the chest, eyes blinking naturally, soft ambient light shifting"
Generate your ideal still frame in Midjourney with precise control over composition and style, then bring it into Runway for animation. This two-step workflow gives you the best of both tools.
In Image to Video mode, your prompt should focus entirely on motion — not re-describing the scene. The image handles the scene; your prompt handles what moves and how.
Subtle, natural motion (gentle breeze, slow rotation, soft breathing) looks more realistic than dramatic movement. Start with subtle motion and increase intensity if needed.
The quality of your output is limited by the quality of your input. Use the highest resolution source image available — at least 1024x1024 pixels for best results.
French
Italian
Spanish
German
Dutch
Arabic
Chinese (Simplified)
Japanese
Korean
Portuguese
Russian
Turkish
Swahili
Hausa
Yoruba
Polish
Sundanese
Kurdish (Kurmanji)