By name: sora2-gen description: Generate videos via OpenAI Sora 2 API (text-to-video, image-to-video, remix). user-invocable: false homepage: https://developers.openai.com/api/docs/guides/video-generation metadata: { "openclaw": { "emoji": "🎥", "requires": { "bins": ["uv"], "env": ["OPENAI_API_KEY"] }, "primaryEnv": "OPENAI_API_KEY", "install": [ { "id": "uv-brew", "kind": "brew", "formula": "uv", "bins": ["uv"], "label": "Install uv (brew)", }, ], },
}
Sora 2 Video Generation
Use the bundled script to generate videos from text prompts, images, or remix existing videos.
Text-to-video
uv run {baseDir}/scripts/generate_video.py --prompt "your video description" --filename "output.mp4"
Portrait video (9:16)
uv run {baseDir}/scripts/generate_video.py --prompt "description" --filename "reel.mp4" --size 720x1280
Square video (1:1)
uv run {baseDir}/scripts/generate_video.py --prompt "description" --filename "square.mp4" --size 1080x1080
Longer clip (8 or 12 seconds)
uv run {baseDir}/scripts/generate_video.py --prompt "description" --filename "long.mp4" --seconds 12
HD quality (Sora 2 Pro)
uv run {baseDir}/scripts/generate_video.py --prompt "description" --filename "hd.mp4" --model sora-2-pro
Image-to-video (animate a reference image)
uv run {baseDir}/scripts/generate_video.py --prompt "she turns and smiles" --filename "animated.mp4" -i first_frame.jpg --size 1280x720
Remix an existing video
uv run {baseDir}/scripts/generate_video.py --remix VIDEO_ID --prompt "shift palette to warm tones" --filename "remix.mp4"
API key
OPENAI_API_KEYenv var- Or set
skills."sora2-gen".apiKey/skills."sora2-gen".env.OPENAI_API_KEYin~/.openclaw/openclaw.json
Parameters
| Flag | Options | Default | Notes |
|---|---|---|---|
--prompt |
text | (required) | Video description |
--filename |
path | (required) | Output .mp4 path |
--size |
1280x720, 720x1280, 1080x1080 |
1280x720 |
Landscape, portrait, or square |
--seconds |
4, 8, 12 |
4 |
Clip length |
--model |
sora-2, sora-2-pro |
sora-2 |
Standard or HD |
--remix |
video ID | — | Remix existing video |
-i |
image path | — | Reference image for image-to-vid |
Model variants
| Model | Quality | Duration | Cost (est.) | Use Case |
|---|---|---|---|---|
sora-2 |
Standard | 4-12s | $0.16-0.48 | Drafts, social, A/B testing |
sora-2-pro |
HD | 4-12s | $0.48-1.44 | Production, cinematic, ads |
Notes
- Duration supports 4, 8, or 12 seconds (longer than Veo 3.1's max of 8s).
- Audio is NOT natively generated — Sora 2 outputs silent video.
- Image input must match the target video resolution (size parameter).
- Remix preserves original structure/composition — best for single focused edits.
- Always use
./relative paths for filenames so OpenClaw can auto-attach via chat (e.g.,./media/generated/drafts/2026-02-09-scene.mp4). - The script prints a
MEDIA:line for OpenClaw to auto-attach on supported chat providers. - Do not read the video back; report the saved path only.
- Latency: 2-5 minutes depending on model, duration, and API load.
- Videos include C2PA metadata and watermarks for AI content identification.
- Content restrictions: no minors in unsafe contexts, no copyrighted characters/music, no real people.