> **Building with AI coding agents?** Install the authstack plugin with one command. This equips your agent with accurate Scalekit implementation patterns.
>
> **Recommended** (the `-y` flags skip prompts, so the command runs without an interactive terminal):
> ```bash
> npx -y @scalekit-inc/cli setup -y
> ```
>
> Global:
> ```bash
> npm install -g @scalekit-inc/cli
> scalekit setup -y
> ```
>
> Supports Claude Code, Cursor, GitHub Copilot, Codex + skills for other Agent Skills-compatible agents in `~/.agents/skills`.
> Skills: integrate-agentkit, implement-saaskit, add-mcp-oauth, implement-sso, implement-scim.
> [Full setup guide](https://docs.scalekit.com/dev-kit/build-with-ai/)

---

# Connect AI agents to Runway

The Runway connector lets your AI agent act in each user's Runway account. Each user connects their own Runway access token once, and Scalekit sends it with every call, so your agent never handles credentials. It comes with 27 tools.

**Authentication:** Bearer Token
**Categories:** AI, Media
**Tools:** 27: 8 read, 17 write, 2 destructive
**Users sign in with:** Access token
**Built by:** Scalekit connector
**Try it:** [Playground in the Scalekit dashboard](https://app.scalekit.com/ws/signup?sk_intent=playground&provider=RUNWAY)

## Setup

1. ### Install the SDK

   **Node.js**

   ```bash
   npm install @scalekit-sdk/node dotenv
   ```

   **Python**

   ```bash
   pip install scalekit-sdk-python python-dotenv
   ```

2. ### Set your credentials

   Add your Scalekit credentials to your `.env` file. Find values in **[app.scalekit.com](https://app.scalekit.com)** > **Developers** > **API Credentials**.

   ```sh title=".env"
   SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
   SCALEKIT_CLIENT_ID=<your-client-id>
   SCALEKIT_CLIENT_SECRET=<your-client-secret>
   ```

3. ### Create the Runway connection

   In **AgentKit > Connections**, create a Runway connection. The name you give it is the `connection_name` your code passes. See [Configure connections](/agentkit/connections/).

4. ### Authorize a user and make your first call

   **Node.js** (`quickstart.mts`)

   ```typescript
   import { ScalekitClient } from '@scalekit-sdk/node'
   import 'dotenv/config'
   import { createInterface } from 'node:readline/promises'

   const scalekit = new ScalekitClient(
     process.env.SCALEKIT_ENVIRONMENT_URL,
     process.env.SCALEKIT_CLIENT_ID,
     process.env.SCALEKIT_CLIENT_SECRET,
   )
   const actions = scalekit.actions

   const connector = 'runway'
   const identifier = 'user_123'

   // Generate an authorization link for the user
   const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
   console.log('Authorize Runway:', link)
   const rl = createInterface({ input: process.stdin, output: process.stdout })
   await rl.question('Press Enter after authorizing...')
   rl.close()

   // Make your first call
   const result = await actions.executeTool({
     connector,
     identifier,
     toolName: 'runway_organization_get',
     toolInput: {},
   })
   console.log(result)
   ```

   ```bash
   npx tsx quickstart.mts
   ```

   **Python** (`quickstart.py`)

   ```python
   import os
   from scalekit import ScalekitClient
   from dotenv import load_dotenv
   load_dotenv()

   scalekit_client = ScalekitClient(
       env_url=os.getenv("SCALEKIT_ENVIRONMENT_URL"),
       client_id=os.getenv("SCALEKIT_CLIENT_ID"),
       client_secret=os.getenv("SCALEKIT_CLIENT_SECRET"),
   )
   actions = scalekit_client.actions

   connection_name = "runway"
   identifier = "user_123"

   # Generate an authorization link for the user
   link_response = actions.get_authorization_link(
       connection_name=connection_name,
       identifier=identifier,
   )
   print("Authorize Runway:", link_response.link)
   input("Press Enter after authorizing...")

   # Make your first call
   result = actions.execute_tool(
       tool_input={},
       tool_name="runway_organization_get",
       connection_name=connection_name,
       identifier=identifier,
   )
   print(result)
   ```

   ```bash
   python quickstart.py
   ```

   Each user opens the link once and enters their Runway credentials there. See [Authorize a user](/agentkit/tools/authorize/) for the full flow and statuses.

## Tools

Pass the exact name to `execute_tool`, with an input like each tool's example. To ask for a tool that's missing, use the [request form](https://scalekitsupport.portal.usepylon.com/forms/request-a-connector-tool).

**Node.js**

```typescript
const result = await actions.executeTool({
  toolName: 'runway_organization_get',
  toolInput: {},
  connector: 'runway',
  identifier: 'user_123',
})
```

**Python**

```python
result = actions.execute_tool(
    tool_name="runway_organization_get",
    tool_input={},
    connection_name="runway",
    identifier="user_123",
)
```

`result.data` is the app's response as JSON, and `result.execution_id` (`executionId` in Node.js) is the ID of the call.

### `runway_organization_get`

Get Organization · Read-only

Get the API organization's credit balance, usage tier and rate limits.
Returns creditBalance, tier (maxMonthlyCreditSpend and per-model limits) and usage (per-model daily generations).
Use this to check credits before starting jobs. Use runway_organization_usage_get for credit usage by day and model.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs: none.

Example input: `{}`

### `runway_organization_usage_get`

Get Credit Usage · Read-only

Get credit usage for the organization broken down by day and model, for up to 90 days.
Returns results (each date with used credits per model) and the list of models.
Use this to report spend. Use runway_organization_get for the current balance.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `beforeDate` (`string`, optional): Day after the last day to include (exclusive), YYYY-MM-DD. Default: today. Example: 2026-11-15
- `startDate` (`string`, optional): First day to include, YYYY-MM-DD. Default: 30 days before beforeDate. Example: 2026-10-15

Example input: `{}`

### `runway_task_get`

Get Task · Read-only

Get the status and output of a Runway generation task.
Returns id, status (PENDING, THROTTLED, RUNNING, SUCCEEDED, FAILED or CANCELLED), createdAt, progress, failure details, and output URLs when it succeeded. Output URLs expire, so download them soon.
Use this after any runway_*_create tool to wait for the result. Poll no more than every 5 seconds.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Task ID returned by any Runway generation tool. Example: 17f20503-6c24-4c16-946b-35dbbce2af2f

Example input: `{"id":"<id>"}`

### `runway_voice_get`

Get Voice · Read-only

Get one custom voice by ID.
Returns id, name, description, status, createdAt and a preview URL when available.
Use this to check whether a new voice is ready. Use runway_voices_list to find IDs.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Custom voice ID. Get it from runway_voices_list or runway_voice_create. Example: 6f4d5f4a-2b1c-4a39-9b55-2c9d0e9a8f31

Example input: `{"id":"<id>"}`

### `runway_voices_list`

List Voices · Read-only

List the organization's custom voices.
Returns data (voices with id, name, description, status and createdAt) and nextCursor.
Use this to find a custom voice ID for runway_text_to_speech_create. Preset voices are not listed; pass them by presetId.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `cursor` (`string`, optional): Cursor from the previous response, to get the next page. Example: eyJpZCI6IjZmNGQ1ZjRhIn0
- `limit` (`integer`, optional): Voices per page, 1 to 100. Default: 50. Example: 50. Default: `50`.

Example input: `{}`

### `runway_workflow_get`

Get Workflow · Read-only

Get a published workflow, including its graph of nodes.
Returns id, name, description and the graph with node IDs and their outputs.
Use this to find the node IDs and output keys to override in runway_workflow_run.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Published workflow ID. Get it from runway_workflows_list or copy it from the Runway developer portal. Example: 0d6c1d1f-7b0a-4b8b-8c41-9a3c5d2e7f10

Example input: `{"id":"<id>"}`

### `runway_workflow_invocation_get`

Get Workflow Invocation · Read-only

Get the status and outputs of a workflow run.
Returns id, status, progress, failure details and the output of each node when finished.
Use this after runway_workflow_run to wait for the result.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Workflow invocation ID returned by runway_workflow_run. Example: 9b2f3c4d-5e6f-4a7b-8c9d-0e1f2a3b4c5d

Example input: `{"id":"<id>"}`

### `runway_workflows_list`

List Workflows · Read-only

List the published workflows the organization can run through the API.
Returns workflows with id, name, description and version details.
Use this to find a workflow ID for runway_workflow_run.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs: none.

Example input: `{}`

### `runway_character_performance_create`

Animate Character Performance · Write

Start a Runway Act-Two job that drives a character image or video with the motion and expressions of a performance video, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for performance capture. Use runway_image_to_video_create for prompt-driven animation.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `character` (`object`, required): Character to animate: {"type": "image", "uri": ...} or {"type": "video", "uri": ...}, where uri is an HTTPS URL, a data URI, or a runway:// URI. Example: {"type": "image", "uri": "https://example.com/images/character.png"}
- `reference` (`object`, required): Driving performance video whose motion and expressions are transferred: {"type": "video", "uri": ...}. Example: {"type": "video", "uri": "https://example.com/videos/performance.mp4"}
- `bodyControl` (`boolean`, optional): If true, transfer body motion as well as facial expression. Example: true
- `expressionIntensity` (`integer`, optional): Strength of facial expressions, 1 to 5. Default: 3. Example: 3
- `ratio` (`string`, optional): Output resolution: 1280:720, 720:1280, 960:960, 1104:832, 832:1104 or 1584:672. Example: 1280:720. One of: `1280:720`, `720:1280`, `960:960`, `1104:832`, `832:1104`, `1584:672`.
- `seed` (`integer`, optional): Random seed, 0 to 4294967295. Reuse a seed with identical inputs to get similar results; leave empty for a random seed. Not every model accepts a seed. Example: 42

Example input: `{"character":{"type":"image","uri":"https://example.com/images/character.png"},"reference":{"type":"video","uri":"https://example.com/videos/performance.mp4"}}`

### `runway_image_to_video_create`

Generate Video from Image · Write

Start a Runway job that animates an image into a video with the chosen model, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this when you have a starting image. Use runway_text_to_video_create for a prompt only and runway_video_to_video_create to edit an existing video.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `model` (`string`, required): Video model to use, for example gen4.5, gen4_turbo, veo3.1, veo3.1_fast, kling3.0_pro or seedance2. Accepted ratio, duration and other fields differ by model. Example: gen4.5. One of: `gen4.5`, `gen4_turbo`, `veo3.1`, `veo3.1_fast`, `klingO3_4k`, `hailuo3`, `kling3.0_4k`, `happyhorse_1_0`, `kling2.5_turbo_pro`, `kling3.0_pro`, `kling3.0_standard`, `klingO3_pro`, `klingO3_standard`, `seedance2`, `seedance2_fast`, `seedance2_mini`, `gemini_omni_flash`, `grok_imagine_1_5`, `wan3`, `gemini_omni_flash_1.1`, `wan3_prime`, `h3_max`, `seedance2_5`, `grok_imagine_1_5_lite`, `kling_4`, `gemini_omni_flash_1.2`, `kling_4_flash`, `hailuo3_highspeed`, `hailuo3_lite`.
- `promptImage` (`string`, required): Image to animate as the first frame: an HTTPS URL, a data URI, or a runway:// URI. Some models also accept an array of {"uri", "position"} objects for first and last frames, which can be passed in options instead. Example: https://example.com/images/storefront.jpg
- `duration` (`integer`, optional): Video length in seconds; required by some models, including gen4.5, and allowed values depend on the model (gen4.5 accepts 2 to 10). Example: 5
- `options` (`object`, optional): Extra model-specific body fields merged into the request as-is, for settings this tool has no input for (for example contentModeration, outputFormat, audio, resolution or references). Field names are Runway's camelCase names and depend on the model; see https://docs.dev.runwayml.com/api/. Example: {"contentModeration": {"publicFigureThreshold": "auto"}}
- `promptText` (`string`, optional): Text prompt describing what should appear in the output. Required by some models. Maximum length depends on the model (1000 characters for the example model). Example: A slow dolly shot through a neon-lit alley in the rain
- `ratio` (`string`, optional): Output resolution as width:height in pixels; required by some models, including gen4.5, and allowed values depend on the model (gen4.5 accepts 1280:720, 720:1280, 1104:832, 960:960, 832:1104 and 1584:672). Example: 1280:720
- `seed` (`integer`, optional): Random seed, 0 to 4294967295. Reuse a seed with identical inputs to get similar results; leave empty for a random seed. Not every model accepts a seed. Example: 42

Example input: `{"model":"gen4.5","promptImage":"<promptImage>"}`

### `runway_image_upscale_create`

Upscale Image · Write

Start a Runway job that upscales an image with the Magnific Precision Upscaler, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this to enlarge an image. Use runway_video_upscale_create for video.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `imageUri` (`string`, required): Image to upscale: an HTTPS URL, a data URI, or a runway:// URI. Example: https://example.com/images/portrait.png
- `flavor` (`string`, optional): Upscaling style: sublime, photo or photo_denoiser. Example: photo. One of: `sublime`, `photo`, `photo_denoiser`.
- `scaleFactor` (`integer`, optional): Upscale factor: 2, 4, 8 or 16. Example: 4. One of: `2`, `4`, `8`, `16`.
- `sharpen` (`integer`, optional): Sharpening strength, 0 to 100. Example: 20
- `smartGrain` (`integer`, optional): Film grain added to hide artifacts, 0 to 100. Example: 10
- `ultraDetail` (`integer`, optional): Amount of fine detail added, 0 to 100. Example: 30

Example input: `{"imageUri":"<imageUri>"}`

### `runway_music_create`

Generate Music · Write

Start a Runway job that composes music from a text description, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for soundtracks. Use runway_sound_effect_create for sound effects.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `promptText` (`string`, required): Description of the music: genre, mood, instruments, tempo, and lyrics if wanted. Up to 4100 characters. Example: Upbeat lo-fi hip hop with warm piano and vinyl crackle
- `duration` (`integer`, optional): Length in seconds, 3 to 600. Default: 30. Example: 60
- `instrumental` (`boolean`, optional): If true, generate music without vocals. Example: true

Example input: `{"promptText":"<promptText>"}`

### `runway_sound_effect_create`

Generate Sound Effect · Write

Start a Runway job that generates a sound effect from a text description, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for effects and ambience. Use runway_music_create for music and runway_text_to_speech_create for speech.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `model` (`string`, required): Sound model: eleven_text_to_sound_v2, mirelo_sfx_2_0 (requires duration) or seed_audio. Example: eleven_text_to_sound_v2. One of: `seed_audio`, `mirelo_sfx_2_0`, `eleven_text_to_sound_v2`.
- `promptText` (`string`, required): Description of the sound. Example: Heavy rain on a tin roof with distant thunder
- `duration` (`number`, optional): Length in seconds. eleven_text_to_sound_v2 accepts 0.5 to 30; mirelo_sfx_2_0 requires it (0.1 to 600); seed_audio does not take it. Example: 5
- `options` (`object`, optional): Extra model-specific body fields merged into the request as-is, for settings this tool has no input for (for example contentModeration, outputFormat, audio, resolution or references). Field names are Runway's camelCase names and depend on the model; see https://docs.dev.runwayml.com/api/. Example: {"loop": true}

Example input: `{"model":"seed_audio","promptText":"<promptText>"}`

### `runway_speech_to_speech_create`

Convert Speech to Another Voice · Write

Start a Runway job that re-speaks the speech in an audio or video file with a different voice, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this to swap a voice and keep the delivery. Use runway_text_to_speech_create to speak new text.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `media` (`object`, required): Audio or video whose speech is re-voiced: {"type": "audio", "uri": ...} or {"type": "video", "uri": ...}, where uri is an HTTPS URL, a data URI, or a runway:// URI. Example: {"type": "audio", "uri": "https://example.com/audio/voiceover.mp3"}
- `voice` (`object`, required): Voice to speak with. Either a Runway preset, {"type": "runway-preset", "presetId": "Maya"} (presets include Maya, Arjun, Serene, Bernard, Leslie, Eleanor, Noah, James, Rachel and others), or a custom voice created with runway_voice_create, {"type": "custom", "id": "<voice id>"}. Example: {"type": "runway-preset", "presetId": "Maya"}
- `removeBackgroundNoise` (`boolean`, optional): If true, remove background noise from the source first. Example: true

Example input: `{"media":{"type":"audio","uri":"https://example.com/audio/voiceover.mp3"},"voice":{"type":"runway-preset","presetId":"Maya"}}`

### `runway_text_to_image_create`

Generate Image · Write

Start a Runway job that generates an image from a prompt and optional reference images, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for still images. Use runway_image_upscale_create to upscale an existing image.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `model` (`string`, required): Image model, for example gen4_image, gen4_image_turbo, gpt_image_2, gemini_image3_pro, seedream5_pro or ideogram_4_5. Accepted ratios and fields differ by model. Example: gen4_image. One of: `gen4_image_turbo`, `gen4_image`, `gpt_image_2`, `gemini_image3.1_flash`, `muse_image`, `seedream5_pro`, `seedream5_lite`, `grok_imagine_image_2`, `gemini_image3.1_flash_lite`, `gemini_image3_pro`, `gpt_image_2_5_flare`, `gpt_image_2_5_sunburst`, `bria_fibo_1_5`, `bria_fibo_edit_1_5`, `ideogram_4_0`, `ideogram_4_5`, `seedream5_flash`, `nano_banana_2.1`, `recraft_v4_1_flash`, `recraft_v4_1_vector`, `ideogram_vector`, `gpt_image_2_6_sunburst`, `gpt_image_2_6_flare`, `seedream5_pro_layers`, `gemini_2.5_flash`.
- `promptText` (`string`, required): Text describing the image. Maximum length depends on the model (1000 characters for gen4_image). Reference images can be named in the prompt with @tag. Example: A product photo of @shoe on a marble pedestal, soft studio light
- `options` (`object`, optional): Extra model-specific body fields merged into the request as-is, for settings this tool has no input for (for example contentModeration, outputFormat, audio, resolution or references). Field names are Runway's camelCase names and depend on the model; see https://docs.dev.runwayml.com/api/. Example: {"contentModeration": {"publicFigureThreshold": "auto"}}
- `ratio` (`string`, optional): Output resolution as width:height in pixels; required by some models, including gen4_image, and allowed values depend on the model (gen4_image accepts 1024:1024, 1920:1080, 1080:1920, 1280:720, 720:1280 and others). Example: 1920:1080
- `referenceImages` (`array`, optional): Reference images to guide the output, each with a uri (an HTTPS URL, a data URI, or a runway:// URI) and an optional tag (up to 16 characters) that the prompt can cite as @tag. gen4_image accepts up to 3. Example: [{"uri": "https://example.com/images/shoe.png", "tag": "shoe"}]
- `seed` (`integer`, optional): Random seed, 0 to 4294967295. Reuse a seed with identical inputs to get similar results; leave empty for a random seed. Not every model accepts a seed. Example: 42

Example input: `{"model":"gen4_image_turbo","promptText":"<promptText>"}`

### `runway_text_to_speech_create`

Generate Speech · Write

Start a Runway job that turns text into speech with a preset or custom voice, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for narration. Use runway_speech_to_speech_create to re-voice existing audio and runway_voice_dubbing_create to translate it.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `model` (`string`, required): Speech model: eleven_v3, eleven_v4, eleven_multilingual_v2 or seed_audio. ElevenLabs models need a voice. Example: eleven_v3. One of: `seed_audio`, `eleven_v4`, `eleven_multilingual_v2`, `eleven_v3`.
- `promptText` (`string`, required): Text to speak. Maximum length depends on the model (eleven_v3: 5000 characters). Example: Welcome back! Your order has shipped and will arrive Tuesday.
- `options` (`object`, optional): Extra model-specific body fields merged into the request as-is, for settings this tool has no input for (for example contentModeration, outputFormat, audio, resolution or references). Field names are Runway's camelCase names and depend on the model; see https://docs.dev.runwayml.com/api/. Example: {"speed": 1.0, "languageCode": "en"}
- `voice` (`object`, optional): Voice to speak with; required for the ElevenLabs models (seed_audio takes a reference-audio voice instead). Either a Runway preset, {"type": "runway-preset", "presetId": "Maya"} (presets include Maya, Arjun, Serene, Bernard, Leslie, Eleanor, Noah, James, Rachel and others), or a custom voice created with runway_voice_create, {"type": "custom", "id": "<voice id>"}. Example: {"type": "runway-preset", "presetId": "Maya"}

Example input: `{"model":"seed_audio","promptText":"<promptText>"}`

### `runway_text_to_video_create`

Generate Video from Text · Write

Start a Runway job that generates a video from a text prompt with the chosen model, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for prompt-only video. Use runway_image_to_video_create to start from an image.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `model` (`string`, required): Video model to use, for example gen4.5, veo3.1, veo3.1_fast, kling3.0_pro or seedance2. Accepted ratio, duration and other fields differ by model. Example: gen4.5. One of: `gen4.5`, `veo3.1`, `veo3.1_fast`, `klingO3_4k`, `hailuo3`, `kling3.0_4k`, `happyhorse_1_0`, `kling2.5_turbo_pro`, `kling3.0_pro`, `kling3.0_standard`, `klingO3_pro`, `klingO3_standard`, `seedance2`, `seedance2_fast`, `seedance2_mini`, `gemini_omni_flash`, `grok_imagine_1_5`, `wan3`, `gemini_omni_flash_1.1`, `wan3_prime`, `h3_max`, `seedance2_5`, `grok_imagine_1_5_lite`, `kling_4`, `gemini_omni_flash_1.2`, `kling_4_flash`, `hailuo3_highspeed`, `hailuo3_lite`.
- `promptText` (`string`, required): Text prompt describing what should appear in the output. Maximum length depends on the model (1000 characters for the example model). Example: A slow dolly shot through a neon-lit alley in the rain
- `duration` (`integer`, optional): Video length in seconds; required by some models, including gen4.5, and allowed values depend on the model (gen4.5 accepts 2 to 10). Example: 5
- `options` (`object`, optional): Extra model-specific body fields merged into the request as-is, for settings this tool has no input for (for example contentModeration, outputFormat, audio, resolution or references). Field names are Runway's camelCase names and depend on the model; see https://docs.dev.runwayml.com/api/. Example: {"contentModeration": {"publicFigureThreshold": "auto"}}
- `ratio` (`string`, optional): Output resolution as width:height in pixels; required by some models, including gen4.5, and allowed values depend on the model (gen4.5 accepts 1280:720 and 720:1280). Example: 1280:720
- `seed` (`integer`, optional): Random seed, 0 to 4294967295. Reuse a seed with identical inputs to get similar results; leave empty for a random seed. Not every model accepts a seed. Example: 42

Example input: `{"model":"gen4.5","promptText":"<promptText>"}`

### `runway_video_frame_rate_enhance`

Enhance Video Frame Rate · Write

Start a Runway job that interpolates a video to a higher or standard frame rate, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this for smoother motion or a delivery frame rate. Use runway_video_upscale_create to raise resolution.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `targetFramerate` (`string`, required): Target frame rate: 24, 25, 30, 48, 50, 60, 120, 23_98, 29_97 or 59_94 (underscores stand for decimals, so 29_97 is 29.97 fps). Example: 60. One of: `24`, `25`, `30`, `48`, `50`, `60`, `120`, `23_98`, `29_97`, `59_94`.
- `videoUri` (`string`, required): Video to convert: an HTTPS URL, a data URI, or a runway:// URI. Example: https://example.com/videos/clip.mp4

Example input: `{"targetFramerate":"24","videoUri":"<videoUri>"}`

### `runway_video_to_video_create`

Edit Video · Write

Start a Runway job that edits or restyles an existing video from a prompt, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this to transform footage you already have. Use runway_text_to_video_create to generate new footage.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `model` (`string`, required): Video-to-video model, for example aleph2, klingO3_pro or seedance2. Accepted fields differ by model. Example: aleph2. One of: `aleph2`, `klingO3_4k`, `hailuo3`, `klingO3_pro`, `klingO3_standard`, `seedance2`, `seedance2_fast`, `seedance2_mini`, `gemini_omni_flash`, `gemini_omni_flash_1.1`, `seedance2_5`, `kling_4`, `gemini_omni_flash_1.2`, `kling_4_flash`, `hailuo3_highspeed`, `hailuo3_lite`.
- `videoUri` (`string`, required): Input video to transform: an HTTPS URL, a data URI, or a runway:// URI. Length limits depend on the model (aleph2: 30 seconds or less). Example: https://example.com/videos/clip.mp4
- `options` (`object`, optional): Extra model-specific body fields merged into the request as-is, for settings this tool has no input for (for example contentModeration, outputFormat, audio, resolution or references). Field names are Runway's camelCase names and depend on the model; see https://docs.dev.runwayml.com/api/. Example: {"targetAspectRatio": "16:9"}
- `promptText` (`string`, optional): Text describing the change to make or what should appear in the output. Example: Turn the scene into a snowy winter evening
- `seed` (`integer`, optional): Random seed, 0 to 4294967295. Reuse a seed with identical inputs to get similar results; leave empty for a random seed. Not every model accepts a seed. Example: 42

Example input: `{"model":"aleph2","videoUri":"<videoUri>"}`

### `runway_video_upscale_create`

Upscale Video · Write

Start a Runway job that upscales a video with the Magnific creative video upscaler, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this to raise resolution. Use runway_video_frame_rate_enhance to change only the frame rate.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `videoUri` (`string`, required): Video to upscale: an HTTPS URL, a data URI, or a runway:// URI. Example: https://example.com/videos/clip.mp4
- `creativity` (`integer`, optional): How much new detail the upscaler may invent, 0 to 100. Example: 20
- `flavor` (`string`, optional): Look of the result: vivid or natural. Example: natural. One of: `vivid`, `natural`.
- `fpsBoost` (`boolean`, optional): If true, also raise the frame rate. Example: false
- `resolution` (`string`, optional): Target resolution: 720p, 1k, 2k or 4k. Example: 4k. One of: `720p`, `1k`, `2k`, `4k`.
- `sharpen` (`integer`, optional): Sharpening strength, 0 to 100. Example: 20
- `smartGrain` (`integer`, optional): Film grain added to hide artifacts, 0 to 100. Example: 10

Example input: `{"videoUri":"<videoUri>"}`

### `runway_voice_create`

Create Voice · Write

Create a custom voice by cloning an audio sample or designing one from a text description.
Returns the voice with id, name and status; it becomes usable when status is READY.
Use this before runway_text_to_speech_create with a custom voice. Use runway_voices_list to find existing voices.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `from` (`object`, required): Source of the voice. Clone from audio: {"type": "audio", "audio": "<uri>"} with a 10-second to 5-minute sample (an HTTPS URL, a data URI, or a runway:// URI). Or design from text: {"type": "text", "prompt": "<description>", "model": "eleven_ttv_v3" or "eleven_multilingual_ttv_v2"}. Example: {"type": "text", "prompt": "A warm, calm, middle-aged British male narrator", "model": "eleven_ttv_v3"}
- `name` (`string`, required): Voice name, up to 100 characters. Example: Brand Narrator
- `description` (`string`, optional): Optional description, up to 1024 characters. Example: Voice for product explainer videos

Example input: `{"from":{"type":"text","prompt":"A warm, calm, middle-aged British male narrator","model":"eleven_ttv_v3"},"name":"<name>"}`

### `runway_voice_dubbing_create`

Dub Audio · Write

Start a Runway job that dubs spoken audio into another language, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this to translate speech. Use runway_speech_to_speech_create to change the voice in the same language.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `audioUri` (`string`, required): Audio to dub: an HTTPS URL, a data URI, or a runway:// URI. Example: https://example.com/audio/interview.mp3
- `targetLang` (`string`, required): Target language code, such as en, es, fr, de, ja, hi, pt or zh. Example: es. One of: `en`, `hi`, `pt`, `zh`, `es`, `fr`, `de`, `ja`, `ar`, `ru`, `ko`, `id`, `it`, `nl`, `tr`, `pl`, `sv`, `fil`, `ms`, `ro`, `uk`, `el`, `cs`, `da`, `fi`, `bg`, `hr`, `sk`, `ta`.
- `disableVoiceCloning` (`boolean`, optional): If true, use a generic voice instead of cloning the original speaker. Example: false
- `dropBackgroundAudio` (`boolean`, optional): If true, leave out the background audio in the output. Example: false
- `numSpeakers` (`integer`, optional): Number of speakers in the audio, if known. Example: 2

Example input: `{"audioUri":"<audioUri>","targetLang":"en"}`

### `runway_voice_isolation_create`

Isolate Voice · Write

Start a Runway job that separates the voice from background noise and music in an audio file, spending credits.
Returns the task id. Generation runs asynchronously; poll runway_task_get with the id until status is SUCCEEDED, then read the output URLs.
Use this to clean up recordings.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `audioUri` (`string`, required): Audio to clean up: an HTTPS URL, a data URI, or a runway:// URI. Example: https://example.com/audio/noisy-call.mp3

Example input: `{"audioUri":"<audioUri>"}`

### `runway_voice_update`

Update Voice · Write

Rename a custom voice or change its description.
Returns the updated voice.
Use this to tidy voice names. Use runway_voice_delete to remove it.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Custom voice ID. Get it from runway_voices_list or runway_voice_create. Example: 6f4d5f4a-2b1c-4a39-9b55-2c9d0e9a8f31
- `description` (`string`, optional): New description, up to 1024 characters. Example: Updated voice for explainer videos
- `name` (`string`, optional): New voice name, up to 100 characters. Example: Brand Narrator v2

Example input: `{"id":"<id>"}`

### `runway_workflow_run`

Run Workflow · Write

Run a published Runway workflow, optionally overriding node outputs such as prompts or input media, spending credits.
Returns the workflow invocation id. Poll runway_workflow_invocation_get with it for the result.
Use this for multi-step pipelines built in Runway. Use the runway_*_create tools for a single model call.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `id` (`string`, required): Published workflow ID. Get it from runway_workflows_list or copy it from the Runway developer portal. Example: 0d6c1d1f-7b0a-4b8b-8c41-9a3c5d2e7f10
- `nodeOutputs` (`object`, optional): Optional overrides for node outputs. Keys are node IDs from runway_workflow_get; each value maps an output key to a typed value. Example: {"5a1c2e3f-4b6d-4e8f-9a0b-1c2d3e4f5a6b": {"prompt": {"type": "primitive", "value": "A red bicycle in the snow"}}}

Example input: `{"id":"<id>"}`

### `runway_task_cancel`

Cancel or Delete Task · Destructive

Cancel a running Runway task, or delete a finished task and its outputs.
Returns no content on success.
Use this to stop a job you no longer need. Use runway_task_get to check its status first.

Read only: no. Destructive: yes. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Task ID returned by any Runway generation tool. Example: 17f20503-6c24-4c16-946b-35dbbce2af2f

Example input: `{"id":"<id>"}`

### `runway_voice_delete`

Delete Voice · Destructive

Delete a custom voice.
Returns no content on success.
Use this to remove a voice you no longer need.

Read only: no. Destructive: yes. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, required): Custom voice ID. Get it from runway_voices_list or runway_voice_create. Example: 6f4d5f4a-2b1c-4a39-9b55-2c9d0e9a8f31

Example input: `{"id":"<id>"}`

## Related

Other AI connectors ([all 216](/agentkit/connectors/?category=ai)):

- [Chorus](/agentkit/connectors/chorus/): Scalekit connector, Username and password, 9 tools
- [Gong](/agentkit/connectors/gong/): Scalekit connector, OAuth, 69 tools
- [Jiminny](/agentkit/connectors/jiminny/): Scalekit connector, Access token, 27 tools


---

## More Scalekit documentation

| Resource | What it contains | When to use it |
|----------|-----------------|----------------|
| [/llms.txt](/llms.txt) | Structured index with routing hints per product area | Start here — find which documentation set covers your topic before loading full content |
| [/llms-full.txt](/llms-full.txt) | Complete documentation for all Scalekit products in one file | Use when you need exhaustive context across multiple products or when the topic spans several areas |
| [sitemap-0.xml](https://docs.scalekit.com/sitemap-0.xml) | Full URL list of every documentation page | Use to discover specific page URLs you can fetch for targeted, page-level answers |
