Skip to content

← The Lab

Freefield: an open-weights Higgsfield-style studio on ComfyUI (120 apps, Modal or your GPU)

Freefield rebuilds a Higgsfield-like image/video/VFX/audio studio as ComfyUI workflows with open weights only: each app is a folder with an app.json and an API-format graph. The README lists about 120 apps, lab GPU runs with time and peak VRAM, and a path to run on Modal or on your own ComfyUI. Around 973 GitHub stars as of this morning.

What it is, per the README. There is no per-app Python UI code: the web app and runner are generated from manifests. Apps include create/edit/inpaint, cinema stills and cinema-video (FLUX 3 still → LTX-2.5 i2v with camera moves), H3 video, genjutsu-style ingredients, reframe, mixed-media, SDR→HDR, camera-copy, upscale and more. Models are files you download (Hugging Face repos listed with size and licence); partner/API nodes are intentionally out. The lab logged dozens of Modal runs (L40S/H100) so every in-app example has a measured time and peak VRAM.

How you run it. Local ComfyUI behind their gateway, or Modal for the heavier graphs (cinema-video and H3 paths peak well above a 24 GB card). MIT licence on the repo. Check each model's own licence (LTX, MiniMax H3, FLUX 3 gated weights, etc.) before client work, and only use footage and likenesses you have rights to.

How it differs from OpenField. We already covered OpenField (Director/Stage/Mirror, MCP tools). Freefield is the bigger app catalogue with a Higgsfield-menu layout and documented lab runs; OpenField leans harder into blocking, phone motion and agent MCP. Pick based on whether you want the menu of apps or the previs/MCP workflow.

Our take: this is the open studio to bookmark if you want a Higgsfield-shaped front end without their closed apps, especially for LTX/H3 experiments on rented GPUs. It is not hot on our signals today (no listed-curator post or GitHub Trending hit in this scout window), but the star count and fresh YouTube explainers mean it is past "quiet repo" status.

Related tools on ReelStack

Skills & repos in our directory

More from the Lab

Topics: ComfyUI · Higgsfield · Open-weight video & audio models