Generation endpoints
Authenticate with your API key as a bearer token in the Authorization header. Both SDKs read it from RUNWAYML_API_SECRET automatically.
/v1/image_to_videoCreate a video generation task. Omit promptImage to generate from text alone on models that support it, such as gen4.5.
Sent as the Authorization header. Your key is used only by your browser for this request — it is never sent to Docsbook or stored.
Model ID, e.g. gen4.5, seedance2_5, gen4_turbo, aleph2
Text description of the video to generate
HTTPS URL, base64 data URI, or runway:// URI for the starting image
Output dimensions, e.g. 1280:720. Model-specific
Output length in seconds. Model-specific ceiling
Fixes the generation for reproducible output
curl -X POST https://api.dev.runwayml.com/v1/image_to_video \
-H "Authorization: Bearer $RUNWAYML_API_SECRET" \
-H "Content-Type: application/json" \
-d '{
"model": "gen4.5",
"promptText": "A serene mountain landscape at sunrise",
"ratio": "1280:720",
"duration": 5
}'{
"id": "17f20503-6c24-4c16-946b-35dbbce2af2f"
}/v1/text_to_imageCreate an image generation task.
Sent as the Authorization header. Your key is used only by your browser for this request — it is never sent to Docsbook or stored.
Model ID, e.g. gen4_image, gen4_image_turbo, gpt_image_2
Text description of the image to generate
Output dimensions, e.g. 1360:768
Reference images. Cost and count limits are model-specific
Number of images to generate. Capped per model
low, medium, high or auto on models with quality tiers
curl -X POST https://api.dev.runwayml.com/v1/text_to_image \
-H "Authorization: Bearer $RUNWAYML_API_SECRET" \
-H "Content-Type: application/json" \
-d '{
"model": "gen4_image",
"promptText": "A beautiful sunset over a calm ocean",
"ratio": "1360:768"
}'/v1/character_performanceDrive a character's performance from a reference video using act_two.
Sent as the Authorization header. Your key is used only by your browser for this request — it is never sent to Docsbook or stored.
Model ID, e.g. act_two
The character to animate, as an image or video
The reference performance video
Output dimensions
/v1/video_upscaleUpscale a video of at most 30 seconds.
Sent as the Authorization header. Your key is used only by your browser for this request — it is never sent to Docsbook or stored.
Must be magnific_video_upscaler_creative
HTTPS URL, data URI or runway:// URI of the source video
720p, 1k, 2k (default) or 4k
Increases output frame rate, which changes frame count and cost
Related#
- Tasks and uploads — retrieving results and uploading files
- Models — valid
modelvalues - Inputs — valid
ratiovalues and asset constraints
Set RUNWAYML_API_SECRET and the examples above run unmodified.