OATDA Video Generation

Generate videos from text descriptions using AI models through OATDA's unified API. Video generation is asynchronous — you submit a request and receive a task ID to poll for status.

When to Use

Use this skill when the user wants to:

• Generate a video from a text description via OATDA
• Create AI-generated video clips
• Use MiniMax, Google Veo, Alibaba Wan, or other video models

Prerequisites

The user needs an OATDA API key. Check in this order:

1. $OATDA_API_KEY environment variable
2. ~/.oatda/credentials.json config file

If neither exists, tell the user:

You need an OATDA API key. Get one at https://oatda.com, then set it:
export OATDA_API_KEY=your_key_here

Step-by-Step Instructions

1. Resolve the API key

# Check env var first; if empty, auto-load from credentials file
if [[ -z "$OATDA_API_KEY" ]]; then
  export OATDA_API_KEY=$(cat ~/.oatda/credentials.json 2>/dev/null | jq -r '.profiles[.defaultProfile].apiKey' 2>/dev/null)
fi

# Verify key exists (show first 8 chars only)
echo "${OATDA_API_KEY:0:8}"

If the output is empty or null, stop and ask the user to configure their API key.

IMPORTANT:

• Never print the full API key. Only show the first 8 characters for verification.
• The key resolution script and subsequent curl commands must run in the same shell session. Each separate bash/terminal invocation starts with an isolated environment where previously exported variables are lost. Either run all commands in one session, or chain them (e.g., export OATDA_API_KEY=... && curl ...).

2. Determine the model

Map common aliases:

User says	Provider	Model
seedance, bytedance (default)	bytedance	seedance-1-5-pro-251215
minimax, t2v	minimax	T2V-01
veo, google veo	google	veo-3.0-generate-preview
wan, alibaba	alibaba	wan-t2v
sora	openai	sora
grok video	xai	grok-2-video

Default: bytedance / seedance-1-5-pro-251215 if no model specified.

3. Discover model-specific parameters

IMPORTANT: Different models support different parameters. Before generating, discover what parameters a model supports:

curl -s -X GET "https://oatda.com/api/v1/llm/models?type=video" \
  -H "Authorization: Bearer $OATDA_API_KEY" | jq '.video_models[] | {id, supported_params}'

This returns each video model's supported_params with:

• type: Parameter type (string, number, boolean, file)
• values: Allowed values for enums
• default: Default value
• description: What the parameter does
• optional: Whether it's required
• accept: For file types, what's accepted (e.g., "image/*")

File-type parameters: Parameters like first_frame_image or last_frame_image require publicly accessible URLs (https://...), not local file paths.

Pass model-specific parameters via the model_params object (see examples below).

4. Make the API call

CRITICAL: The endpoint URL includes ?async=true query parameter.

curl -s -X POST "https://oatda.com/api/v1/llm/generate-video?async=true" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OATDA_API_KEY" \
  -d '{
    "provider": "<PROVIDER>",
    "model": "<MODEL>",
    "prompt": "<VIDEO_DESCRIPTION>"
  }'

Replace <PROVIDER>, <MODEL>, and <VIDEO_DESCRIPTION> with actual values.

Optional parameters (add to body):

• duration: Video duration in seconds
• resolution: "720P" or "1080P"
• aspectRatio: "16:9", "9:16", or "1:1"
• quality: Quality setting (model-dependent)
• style: Style setting (model-dependent)
• width / height: Explicit pixel dimensions
• model_params: Model-specific parameters as key-value pairs. Use list_models?type=video or /api/v1/llm/models to discover supported params per model. Examples:
  • Seedance: { "ratio": "16:9", "duration": "5", "generate_audio": true, "camera_fixed": false }
  • Seedance I2V: { "first_frame_image": "https://...", "last_frame_image": "https://..." }
  • MiniMax: { "first_frame_image": "https://...", "resolution": "720P" }
  • xAI: { "resolution": "720p" }

5. Parse the response

{
  "taskId": "minimax-T2V01-abc123def456",
  "status": "pending",
  "provider": "minimax",
  "model": "T2V-01",
  "message": "Video generation task created",
  "pollUrl": "/api/v1/llm/video-status/minimax-T2V01-abc123def456"
}

• Note the taskId — this is needed to check status later
• The initial status will be "pending" or "processing"

6. Tell the user the next step

After submitting, inform the user:

Video generation started! Task ID: <taskId>
Video generation typically takes 30 seconds to 5 minutes.
Use /oatda:oatda-video-status <taskId> to check when your video is ready.

7. Handle errors

HTTP Status	Meaning	Action
401	Invalid API key	Tell user to check their key
400	Bad request / prompt too long	Keep prompt under 4000 chars
429	Rate limited	Wait and retry
400 with content_policy	Content policy violation	Ask user to adjust description

Full Examples

Bytedance Seedance (default)

User asks: "Generate a 5-second video of ocean waves at sunset"

curl -s -X POST "https://oatda.com/api/v1/llm/generate-video?async=true" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OATDA_API_KEY" \
  -d '{
    "provider": "bytedance",
    "model": "seedance-1-5-pro-251215",
    "prompt": "Ocean waves crashing on a beach at sunset, golden hour lighting, cinematic",
    "model_params": {
      "ratio": "16:9",
      "duration": "5",
      "generate_audio": true,
      "camera_fixed": false
    }
  }'

Seedance Image-to-Video (with first and last frame)

User asks: "Create a video transition between these two images"

curl -s -X POST "https://oatda.com/api/v1/llm/generate-video?async=true" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OATDA_API_KEY" \
  -d '{
    "provider": "bytedance",
    "model": "seedance-1-5-pro-251215",
    "prompt": "Smooth transition from day to night",
    "model_params": {
      "ratio": "16:9",
      "first_frame_image": "https://example.com/daytime.jpg",
      "last_frame_image": "https://example.com/nighttime.jpg"
    }
  }'

MiniMax with reference image

User asks: "Animate this image with MiniMax"

curl -s -X POST "https://oatda.com/api/v1/llm/generate-video?async=true" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OATDA_API_KEY" \
  -d '{
    "provider": "minimax",
    "model": "T2V-01",
    "prompt": "The character starts walking forward slowly",
    "model_params": {
      "first_frame_image": "https://example.com/character.png",
      "resolution": "720P"
    }
  }'

Google Veo

User asks: "Generate a cinematic video with Google Veo"

curl -s -X POST "https://oatda.com/api/v1/llm/generate-video?async=true" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OATDA_API_KEY" \
  -d '{
    "provider": "google",
    "model": "veo-3.0-generate-preview",
    "prompt": "A drone shot flying over a misty mountain range at sunrise",
    "duration": 5,
    "aspectRatio": "16:9"
  }'

xAI Grok Video

User asks: "Generate a video with xAI Grok"

curl -s -X POST "https://oatda.com/api/v1/llm/generate-video?async=true" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OATDA_API_KEY" \
  -d '{
    "provider": "xai",
    "model": "grok-2-video",
    "prompt": "A futuristic city with flying cars and holographic billboards",
    "model_params": {
      "resolution": "720p"
    }
  }'

Tips

• Always use ?async=true in the URL — the API does not support synchronous video generation
• Video generation takes 30 seconds to 5+ minutes depending on model and complexity
• Always give the user the task ID and suggest /oatda:oatda-video-status to check progress
• Use list_models?type=video to discover model-specific parameters before generating
• Use model_params for model-specific options (ratio, generate_audio, camera_fixed, etc.)
• For image-to-video: Provide first_frame_image and optionally last_frame_image as public URLs
• Prompt maximum is 4000 characters
• NEVER expose the full API key in output
• Related skills: /oatda:oatda-video-status (required companion for checking results), /oatda:oatda-list-models for available video models