ContextQMD
Back to Openclaw

Runway

docs/providers/runway.md

2026.5.52.5 KB
Original Source

OpenClaw ships a bundled runway provider for hosted video generation.

PropertyValue
Provider idrunway
AuthRUNWAYML_API_SECRET (canonical) or RUNWAY_API_KEY
APIRunway task-based video generation (GET /v1/tasks/{id} polling)

Getting started

<Steps> <Step title="Set the API key"> ```bash openclaw onboard --auth-choice runway-api-key ``` </Step> <Step title="Set Runway as the default video provider"> ```bash openclaw config set agents.defaults.videoGenerationModel.primary "runway/gen4.5" ``` </Step> <Step title="Generate a video"> Ask the agent to generate a video. Runway will be used automatically. </Step> </Steps>

Supported modes

ModeModelReference input
Text-to-videogen4.5 (default)None
Image-to-videogen4.51 local or remote image
Video-to-videogen4_aleph1 local or remote video
<Note> Local image and video references are supported via data URIs. Text-only runs currently expose `16:9` and `9:16` aspect ratios. </Note> <Warning> Video-to-video currently requires `runway/gen4_aleph` specifically. </Warning>

Configuration

json5
{
  agents: {
    defaults: {
      videoGenerationModel: {
        primary: "runway/gen4.5",
      },
    },
  },
}

Advanced configuration

<AccordionGroup> <Accordion title="Environment variable aliases"> OpenClaw recognizes both `RUNWAYML_API_SECRET` (canonical) and `RUNWAY_API_KEY`. Either variable will authenticate the Runway provider. </Accordion> <Accordion title="Task polling"> Runway uses a task-based API. After submitting a generation request, OpenClaw polls `GET /v1/tasks/{id}` until the video is ready. No additional configuration is needed for the polling behavior. </Accordion> </AccordionGroup> <CardGroup cols={2}> <Card title="Video generation" href="/tools/video-generation" icon="video"> Shared tool parameters, provider selection, and async behavior. </Card> <Card title="Configuration reference" href="/gateway/config-agents#agent-defaults" icon="gear"> Agent default settings including video generation model. </Card> </CardGroup>