RunwayML Aleph 2

Third-party

RunwayML's video editing model. Edit one frame to update your whole video, make changes across multiple shots, and work with up to 30 seconds of video. Supports keyframe-guided editing for precise control over specific moments in the clip.

Model Info
Terms and License	link ↗
More information	link ↗
Pricing	View pricing in the Cloudflare dashboard ↗

Usage

const response = await env.AI.run(
  'runwayml/aleph-2',
  {
    prompt: 'Transform the scene into a golden-hour sunset with warm lighting',
    video_uri: 'https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/cinematic-scene.mp4',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "runwayml/aleph-2",
  "input": {
    "prompt": "Transform the scene into a golden-hour sunset with warm lighting",
    "video_uri": "https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/cinematic-scene.mp4"
  }
}'

{
  "state": "Completed",
  "result": {
    "video": "https://dnznrvs05pmza.cloudfront.net/7449093b-e574-41a5-8308-05e4692fa652.mp4"
  },
  "gatewayMetadata": {
    "keySource": "Unified"
  }
}

Examples

Keyframe-Guided Edit — Edit a video using a reference image anchored to the first frame of the output

const response = await env.AI.run(
  'runwayml/aleph-2',
  {
    prompt: 'Edit the video to match the provided reference frame',
    video_uri: 'https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/nature-close-up.mp4',
    prompt_images: [
      {
        uri: 'https://upload.wikimedia.org/wikipedia/commons/8/85/Tour_Eiffel_Wikimedia_Commons_(cropped).jpg',
        position: 'first',
      },
    ],
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "runwayml/aleph-2",
  "input": {
    "prompt": "Edit the video to match the provided reference frame",
    "video_uri": "https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/nature-close-up.mp4",
    "prompt_images": [
      {
        "uri": "https://upload.wikimedia.org/wikipedia/commons/8/85/Tour_Eiffel_Wikimedia_Commons_(cropped).jpg",
        "position": "first"
      }
    ]
  }
}'

{
  "state": "Completed",
  "result": {
    "video": "https://dnznrvs05pmza.cloudfront.net/cd7388fe-09d6-4b16-afc5-24bb315eaf14.mp4"
  },
  "gatewayMetadata": {
    "keySource": "Unified"
  }
}

Timed Keyframe Edit — Place a guidance image at a specific timestamp within the input video using keyframes[].seconds

const response = await env.AI.run(
  'runwayml/aleph-2',
  {
    prompt: 'Change the background to a futuristic cityscape at night',
    video_uri: 'https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/simple-text-to-video.mp4',
    keyframes: [
      {
        uri: 'https://upload.wikimedia.org/wikipedia/commons/8/85/Tour_Eiffel_Wikimedia_Commons_(cropped).jpg',
        seconds: 2.0,
      },
    ],
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "runwayml/aleph-2",
  "input": {
    "prompt": "Change the background to a futuristic cityscape at night",
    "video_uri": "https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/simple-text-to-video.mp4",
    "keyframes": [
      {
        "uri": "https://upload.wikimedia.org/wikipedia/commons/8/85/Tour_Eiffel_Wikimedia_Commons_(cropped).jpg",
        "seconds": 2.0
      }
    ]
  }
}'

{
  "state": "Completed",
  "result": {
    "video": "https://dnznrvs05pmza.cloudfront.net/a7e054fa-1cf2-4265-a35c-518b8c4d6a1a.mp4"
  },
  "gatewayMetadata": {
    "keySource": "Unified"
  }
}

Content Moderation Override — Edit a video featuring public figures with relaxed content moderation

const response = await env.AI.run(
  'runwayml/aleph-2',
  {
    prompt: 'Make the background a dramatic stormy sky',
    video_uri: 'https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/cinematic-scene.mp4',
    content_moderation: {
      public_figure_threshold: 'low',
    },
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "runwayml/aleph-2",
  "input": {
    "prompt": "Make the background a dramatic stormy sky",
    "video_uri": "https://pub-04a6d208d361438ea01b797e6973bd19.r2.dev/catalog/runwayml__gen-4.5/cinematic-scene.mp4",
    "content_moderation": {
      "public_figure_threshold": "low"
    }
  }
}'

{
  "state": "Completed",
  "result": {
    "video": "https://dnznrvs05pmza.cloudfront.net/ea5b1ccd-3dd6-43c3-86e3-5ed720849e92.mp4"
  },
  "gatewayMetadata": {
    "keySource": "Unified"
  }
}

Parameters

prompt

stringrequiredminLength: 1maxLength: 1000Text prompt describing the edit to apply to the input video

video_uri

stringrequiredHTTPS URL, Runway URI, or data URI of the source video to edit (≤30 seconds)

▶keyframes[]

arrayminItems: 1maxItems: 5Timed guidance images placed at specific points in the input video. Each entry has a uri and either seconds (absolute timestamp) or at (fractional position). Up to 5.

▶prompt_images[]

arrayminItems: 1maxItems: 5Image keyframes for guiding the edit at specific points in the output video. Up to 5.

seed

integerminimum: 0maximum: 4294967295Random seed for reproducible results

▶content_moderation{}

objectSettings that affect the behavior of the content moderation system

duration

numberexclusiveMinimum: 0Duration of the source video in seconds. Used for billing only — not sent to RunwayML. Provide this so the gateway can compute per-second video cost accurately.

video

stringformat: uriURL to the edited video

API Schemas (Raw)

Input

Output