Video Generator (Legacy) Block

The Video Generator block creates AI videos from text prompts and optional start/end images using multiple providers (Fal.ai, Runway, Luma, Minimax, and Google Veo). This is a legacy block — reach for individual provider blocks when you need finer control, but it remains handy for quickly switching providers behind a single configuration.

Overview

When to Use

• Generate a short video clip from a text description (text-to-video).
• Animate a still image into a video by supplying a start image (image-to-video).
• Interpolate between a start and end image using Luma, Runway, or Fal.ai.
• Compare or swap between video providers (Fal.ai, Runway, Luma, Minimax, Veo) without rewiring your workflow.
• Produce videos with specific duration, resolution, and aspect-ratio requirements.
• Add reproducibility with a fixed seed, or steer output away from unwanted elements with a negative prompt.

Configuration

Provider (provider)

Required. Choose the video generation backend. The selected provider determines which model dropdown is shown.

Fal.ai Model (falaiModel)

Shown when Provider is falai. Default: fal-ai/wan/v2.1/1.3b/text-to-video.

Runway Model (runwayModel)

Shown when Provider is runway. Default: gen4_turbo.

Luma Model (lumaModel)

Shown when Provider is luma. Default: ray-2-flash.

Minimax Model (minimaxModel)

Shown when Provider is minimax. Default: video-01.

Veo Model (veoModel)

Shown when Provider is veo. Default: veo-2.

Prompt (prompt)

Required. Free-text description of the video you want to generate (long-input field).

Duration (duration)

Optional. Clip length in seconds (e.g. 5). Supported values depend on the provider — Runway accepts 5 or 10 seconds.

Resolution (resolution)

Optional. Output resolution. Default: provider default.

Aspect Ratio (aspectRatio)

Optional. Output aspect ratio. Default: 16:9.

Start Image (imageFile / imageUrl)

Optional. Provide a start image for image-to-video generation either by uploading a file (imageFile) or entering a URL (imageUrl). When both are provided, imageFile takes precedence. Available in basic/advanced mode respectively.

End Image URL (endImageUrl) — Advanced

Optional. URL of an image to use as the last frame for interpolation. Available when Provider is luma, runway, or falai.

Enable Audio (enableAudio) — Advanced

Optional toggle. Enables audio generation. Available when Provider is minimax or falai.

Camera Motion (cameraMotion) — Advanced

Optional. Camera motion description string (e.g. camera dolly forward). Available when Provider is luma.

Negative Prompt (negativePrompt) — Advanced

Optional. Elements to exclude from the generated video.

Seed (seed) — Advanced

Optional. Integer seed for reproducible results.

API Key (apiKey)

Required. Provider-specific API key (stored as a password field). Use {{YOUR_PROVIDER_API_KEY}} to reference a workflow secret.

Inputs & Outputs

• Inputs:

  • provider (string) — Video generation provider
  • falaiModel (string) — Fal.ai model ID
  • runwayModel (string) — Runway model ID
  • lumaModel (string) — Luma model ID
  • minimaxModel (string) — Minimax model ID
  • veoModel (string) — Veo model ID
  • prompt (string) — Video generation prompt
  • duration (string) — Video duration in seconds
  • resolution (string) — Output resolution
  • aspectRatio (string) — Output aspect ratio
  • imageUrl (string) — Start image URL
  • imageFile (json) — Start image file object
  • endImageUrl (string) — End image URL for interpolation
  • enableAudio (boolean) — Whether to enable audio generation
  • cameraMotion (string) — Camera motion description (Luma)
  • negativePrompt (string) — Elements to exclude from the video
  • seed (string) — Random seed for reproducibility
  • apiKey (string) — Provider API key

• Outputs:

  • videoUrl (string) — URL to the generated video
  • id (string) — Generation job ID
  • status (string) — Generation status
  • duration (number) — Video duration in seconds
  • thumbnailUrl (string) — Video thumbnail URL

Tools

• Generate Video with Runway (runway_generate_video) — Generates a video via the Runway API (Gen4 Turbo or Gen3 Alpha Turbo). Accepts a text prompt and optional start/end frame images, duration, aspect ratio, negative prompt, and seed.
• Generate Video with Luma AI (luma_generate_video) — Generates a video using Luma AI Dream Machine models (Ray 2 / Ray 2 Flash / Ray 1.6 Fast). Supports image-to-video, camera motion control, resolution, aspect ratio, negative prompt, and seed.
• Generate Video with MiniMax (minimax_generate_video) — Generates a video using MiniMax Hailuo video models. Accepts prompt, duration, resolution, start image, prompt optimizer toggle, negative prompt, and seed.
• Generate Video with fal.ai (falai_generate_video) — Generates a video using any fal.ai video endpoint (Wan, Kling, Seedance, LTX, etc.). Accepts prompt, model endpoint, duration, aspect ratio, start image, audio flag, negative prompt, and seed.
• Generate Video with Google Veo (veo_generate_video) — Generates a video using Google Veo models (Veo 2 or Veo 3 Fast). Accepts prompt, model, duration, aspect ratio, resolution, start image, audio flag, negative prompt, and seed.

YAML Example

video_gen_1:
  type: video_generator
  name: "Video Generator (Legacy)"
  inputs:
    provider: "falai"
    falaiModel: "fal-ai/wan/v2.1/1.3b/text-to-video"
    prompt: "A serene mountain lake at sunrise, gentle ripples on the water"
    duration: "5"
    resolution: "720p"
    aspectRatio: "16:9"
    apiKey: "{{FALAI_API_KEY}}"
  connections:
    outgoing:
      - target: next-block-id