GTtoken HappyHorse I2V

Purpose

Use this skill to generate image-to-video clips through the user's GTtoken/OpenAI-compatible relay.

Default target:

• Base URL: https://api.modagent-homing.com/v1
• Model: happyhorse-1.0-i2v
• Endpoint style: OpenAI-compatible chat/completions with image input.

This skill is for video generation only. Keep OpenClaw's normal Models page on a working chat model such as deepseek-v4-flash; do not set happyhorse-1.0-i2v as the default chat model.

Required environment variables

Set one API key variable before using the script:

export GTTOKEN_API_KEY="sk-..."

Optional overrides:

export GTTOKEN_BASE_URL="https://api.modagent-homing.com/v1"
export HAPPYHORSE_I2V_MODEL="happyhorse-1.0-i2v"

HAPPYHORSE_API_KEY is accepted as a fallback if GTTOKEN_API_KEY is missing.

Workflow

1. Use this skill only when the request includes an image, image URL, or a recent uploaded image attachment.
2. Extract or ask for the image URL/path from the OpenClaw/WeChat attachment context.
3. Call scripts/generate_happyhorse_i2v.py with --prompt plus either --image-url or --image-file.
4. Return the generated video URL or downloaded output path if the relay returns one.
5. If the relay returns a task id but no final URL, show the raw JSON and explain that the GTtoken relay requires a polling/result endpoint from its documentation.

Quick commands

Using an image URL:

python scripts/generate_happyhorse_i2v.py \
  --image-url "https://example.com/input.jpg" \
  --prompt "把这张图生成5秒视频，镜头缓慢推进，电影感，4K画质" \
  --duration 5 \
  --aspect-ratio 16:9 \
  --output-dir /tmp/happyhorse-i2v

Using a local image file:

python scripts/generate_happyhorse_i2v.py \
  --image-file /path/to/input.jpg \
  --prompt "人物自然微笑，头发轻微飘动，镜头缓慢推进，画面稳定" \
  --duration 5 \
  --output-dir /tmp/happyhorse-i2v

Prompt rules

For best results, include:

• what should move in the original image
• camera movement
• duration, usually 3-15 seconds
• aspect ratio: 16:9, 9:16, 1:1, 4:3, or 3:4
• whether sound or no sound is desired

Prefer one clear visual beat, for example: "镜头缓慢推进，主体保持一致，背景轻微运动，电影感光影".

Troubleshooting

• 401 or invalid api key: the GTtoken key is wrong, expired, or not authorized for HappyHorse.
• model not found: the exact model id is wrong or unavailable to this key. Use happyhorse-1.0-i2v from the GTtoken model page.
• HTML / response: the base URL is wrong; use https://api.modagent-homing.com/v1, not /console/token or the website homepage.
• bad_response_body when testing in OpenClaw Models: expected for video models. Use this skill instead of the Models connectivity test.
• Unsupported image input: the relay may use a nonstandard image field. Retry the script with --payload-mode simple or --payload-mode image_url_field.