Models / Runway / Image to video

Runway image to video prompting guide

How to prompt Runway in image-to-video (start frame) mode. Every point links its source; community tips are marked. Your agent gets the same guide through the Atlas MCP (get_guide).

Prompt structure

camera motion→subject action→additional

• T2V: '[Camera] shot of [subject] [action] in [environment]. [Supporting descriptions]'. Structure matters less than removing ambiguity. source ↗

What the text should describe

• I2V: 'The camera [motion] as the subject [action]. [Additional descriptions]'. Focus almost entirely on motion; describe visuals only for new elements. source ↗
• Refer to characters/objects in the image with general language ('the subject', 'the woman') to isolate them and define their motion. source ↗

What not to describe

• Input images with implied motion (blur, dust, mid-action poses) fight prompts asking for stillness; artifacts in the input get worse. source ↗

Camera vocabulary

• For still shots: 'The locked-off camera remains perfectly still.' source ↗

Responds well to

• Sequential prompting: 'X, then Y. Finally, Z.' or '[00:01] X. [00:03] Y.' source ↗

Duration and limits

• Chain longer sequences: last frame of one generation as the next image input. source ↗

Avoid and failure modes

• Positive, concrete phrasing; no conflicting instructions; very long multi-paragraph prompts over-constrain. source ↗
• Unwanted cuts = duration too short; add 'Continuous, seamless shot'. source ↗

Negative prompt

Supported via .

Real prompts, credited

A chameleon changing color naturally, with its skin shifting through vibrant hues of blue, green, and orange. The camera slowly zooms in on the chameleon as its colors change, highlighting the intricate

Next

Sources

• academy.runwayml.com/guides/prompting-guide
• docs.dev.runwayml.com/guides/models/