Corespeed Art — Multi-Model AI Media via fal.ai
Auth: Set FAL_KEY with your fal.ai API key (get one at https://fal.ai/dashboard/keys).
Workflow
- Pick a model from the tables below
- Read its reference file to get the exact endpoint and parameters
- Run the command with the endpoint and JSON parameters
Usage
uv run {baseDir}/scripts/fal.py ENDPOINT --json '{"param":"value"}' -f output.ext [-i input.ext]
ENDPOINT— the fal.ai model path from the reference file (e.g.fal-ai/nano-banana-2)--json— model parameters as JSON object-f— output filename-i— input file(s) to upload (repeat for multiple), auto-injected asimage_url/image_urls/start_image_url/video_url--audio— audio input file (for lipsync)
Image Generation
| Model | Best For | Reference |
|---|---|---|
| Nano Banana 2 | Pro quality, web search, thinking | Read nanobanana.md |
| FLUX 2 Pro | Photorealistic, zero-config | Read flux.md |
| FLUX Schnell | ⚡ Fastest iteration | Read flux.md |
| FLUX Pro v1.1 | Accelerated, commercial use | Read flux.md |
| FLUX.1 Dev | 12B params, fine-tuning friendly | Read flux.md |
| GPT Image 1.5 | Transparent bg, instruction following | Read gpt.md |
| Qwen Image 2 Pro | Chinese+English, typography, native 2K | Read qwen.md |
| Recraft V4 Pro | Design/marketing, color control | Read recraft.md |
| Seedream 5 Lite | Multi-image editing, reasoning | Read seedream.md |
Video Generation
| Model | Best For | Reference |
|---|---|---|
| Kling v3 Pro I2V | Best I2V, multi-shot, audio, 3–15s | Read kling.md |
| Sora 2 T2V | Long video up to 20s, characters | Read sora.md |
| Sora 2 I2V | Image→video with Sora | Read sora.md |
| Veo 3.1 T2V | Cinematic + native audio/dialogue | Read veo.md |
| Veo 3.1 I2V | Image→video with audio | Read veo.md |
| LTX 2.3 T2V Fast | ⚡ Fast, up to 2160p/20s, open source | Read ltx.md |
| LTX 2.3 I2V | Image→video, start+end frame | Read ltx.md |
| Pixverse v5 I2V | Anime, 3D, clay, cyberpunk styles | Read pixverse.md |
Audio / TTS
| Model | Best For | Reference |
|---|---|---|
| MiniMax Speech-02 HD | 30+ languages, loudness normalization | Read minimax-speech.md |
Music & Sound Effects
| Model | Best For | Reference |
|---|---|---|
| Beatoven Music | AI music, up to 90s | Read beatoven-music.md |
Utilities
| Tool | Best For | Reference |
|---|---|---|
| Topaz Upscale | AI image/video upscale 2x–4x | Read topaz.md |
| BRIA RMBG | Professional background removal | Read bria-rmbg.md |
| Sync Lipsync | Audio-driven lip sync on video | Read sync-lipsync.md |
Notes
- No manual Python setup required. The script uses PEP 723 inline metadata.
uv runautomatically creates an isolated virtual environment and installs thefal-clientdependency on first run. - fal.ai uses a queue system — the script polls until generation completes.
- Video generation can take 30s–3min.
- Use timestamps in filenames:
yyyy-mm-dd-hh-mm-ss-name.ext. - Script prints
MEDIA:line for OpenClaw to auto-attach. - Do not read generated media back; report the saved path only.