LLM (Groq, open model)
Expands your one sentence into a 4–6 shot list with camera, lens and mood. Any OpenAI-compatible endpoint; open-weight models on Groq by default. No key? A deterministic mock writes the shots.
Comfy SDK
@comfyorg/sdk submits our exported ComfyUI workflows (in /workflows) and polls each job.
Comfy Cloud
Runs the workflows on cloud GPUs: Z-Image-Turbo paints each shot's first frame.
LTX-2.3
Image-to-video: animates every first frame into a 2–4s shot, with generated audio.
ffmpeg.wasm
Right in your browser: title card, frame-exact padding, crossfades and the final MP4. No server-side video processing.
Open source · #OpenSourceHiggsfield
Press Enter to write the shot list.
No shots yet. One sentence is all it takes — try one of these:
