Menu
Generate video from audio + image using fal.ai LTX-2.3. Use for: talking head, lip sync, audio-driven video.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Create multi-camera AI music videos from audio. Use for: music video, concert video, multi-angle video, AI video clip.
Beautiful animated lyrics overlay for videos using Remotion. Use for: lyrics, captions, subtitles, karaoke, song text overlay.
Transcribe audio/video to SRT subtitles using ElevenLabs Scribe v2. Use for: transcription, subtitles, captions, SRT generation.
Transcribe audio/video to SRT subtitles using ElevenLabs Scribe v2. Use for: transcription, subtitles, captions, SRT generation.
Generate images with Gemini (default) or fal.ai FLUX.2 klein 4B (--cheap for fast/low-cost). Use for: create image, generate visual, AI image generation, poster.
Render Remotion video compositions to MP4/WebM. Use for: rendering videos, exporting compositions, batch renders, previewing.
| name | audio-to-video |
| description | Generate video from audio + image using fal.ai LTX-2.3. Use for: talking head, lip sync, audio-driven video. |
| allowed-tools | Bash, Read, Write |
Generate video from audio using fal.ai LTX-2.3.
cd /Users/aviz/.claude/skills/audio-to-video/scripts
npx ts-node generate.ts \
--audio "/path/to/audio.mp3" \
--image "/path/to/frame.jpg" \
-d /tmp/output.mp4 \
"Cinematic description of motion and action"
| Flag | Description |
|---|---|
--audio, -a | Audio file (mp3/wav/ogg/m4a/aac) — max 20 seconds |
-d, --destination | Output video path (required) |
--image, -i | Starting frame image (highly recommended) |
--end-image | Ending frame image |
--size, -s | Video size (see below) |
--fps | Frames per second (24/25) |
--quality | low/medium/high/maximum |
--guidance-scale | Default: 9 with image, 5 without |
| Mode | Endpoint |
|---|---|
| Audio + (optional image) | fal-ai/ltx-2.3/audio-to-video |
| Image only (no audio) | fal-ai/ltx-2.3/image-to-video |
Script auto-selects based on whether --audio is provided.
landscape_16_9 (default), portrait_16_9, landscape_4_3, square_hd, auto
Max 20 seconds per clip. For longer content, split into chunks and merge with ffmpeg.
Uses FAL_KEY from ~/.claude/skills/image-generation/scripts/.env
The prompt drives video motion. Include:
Example:
"LIVE ARENA CONCERT: Female singer throwing head back mid-high-note, hair arcing in slow motion.
Camera: slow push in. Lighting: single white spotlight with rim halo.
Fog at feet, pure emotional catharsis. Hyper-real 4K cinema"