utensils/mold

Local AI image and video generation on your own GPU. Mold supports NVIDIA CUDA
and Apple Silicon Metal, with a CLI, native desktop app, web studio, mobile
companions, Discord bot, and REST/SSE API built on the same engine.
**Documentation** ·
**Models** ·
**Desktop guide** ·
**API**
The generation composer offers **Edit prompt** for spacious long-text editing,
recent-prompt search and Clear with Undo. Native iOS and macOS Library Settings
also let you hide date separators for a continuous grid.
Install
Stable release:
Nightly CLI from the latest published main build:
Automatic cloud provisioning uses a verified container digest for the client’s exact
source revision. Nightly/source builds wait for their matching container readiness
manifest; an unpublished image fails before creating a paid host.
The installer selects a compatible build and verifies its checksum. Linux x86_64
clients without a visible NVIDIA GPU receive a GPU-free CLI. To use only remote
GPU hosts, including from a machine with NVIDIA hardware:
The CPU archive needs no NVIDIA driver or CUDA libraries. Local GPU generation
requires a CUDA build (or Metal on macOS). See the
installation guide for Nix,
Arch, Windows, Android, and source builds.
GH200, GB200, and GB300 require future linux/arm64 artifacts and are unsupported.
Mold no longer publishes new versions to crates.io. Existing registry versions
are historical; use GitHub releases, Nix, Docker, AUR, or a source build
for current versions.
Quick start
For API-only servers, set web_ui_enabled = false in config.toml (or
MOLD_WEB_UI_ENABLED=false) and restart. The interface defaults on; disabling
it leaves every API route available. NixOS: services.mold.webUi.enable = false;.
Models download automatically on first use. Generated media is saved locally
with prompt, model, seed, and generation metadata.
LoRA reuse tracks adapter file revisions as well as strengths; replacing an
adapter invalidates cached weights. FLUX and SD3 retain at most 256 MiB of
derived LoRA deltas per engine and release that cache when unloaded.
Web, desktop and mobile reject empty video/audio attachments and files above
64 MiB before encoding, with inline errors that preserve the prior attachment.
Mobile also enforces a 45 MiB combined media budget.
Framewise video upscale also needs the host codec bridge: Nix packages and
CUDA containers include it, while raw binary installs must provide ffmpeg
and ffprobe on PATH before the server advertises that feature.
What it supports
- **Models:** FLUX.1, Flux.2, Stable Diffusion, Z-Image, Qwen-Image,
Qwen Image 2.1, Wuerstchen, LTX Video, Wan, MiniMax H3, and Hunyuan3D. See the
model catalog for variants and hardware
requirements.
Wan's 1.3B BF16 and 5B Q8/FP16 paths are performance-qualified on Apple
Metal as well as CUDA; fp8-scaled Wan checkpoints remain CUDA-only.
MiniMax H3's compact runtime is supported on SM89 CUDA and Apple Metal;
Metal admission checks each request against live unified-memory headroom.
FL2VA accepts text-only (no frame), first, last, or both boundary frames,
with roughly 1,000 prompt tokens with paired frames and charging their
additional memory. Distilled FL2VA tiers include Turbo 4/8-step, DMAD
4-step, PDMD 2/4-step, and FastVideo's FastH3 V2 8-step (CUDA-qualified).
Queued FL2VA jobs preserve first/last endpoint roles through durable storage
and replay, including all supported Turbo tiers.
Forced-local H3 execution accepts one FL2VA request; batches, sequences, and
Ref2VA reference uploads require the server route.
Remote Ref2VA automatically selects authenticated uploads or bounded inline
references on hosts with authentication disabled; no client
API key is needed for the latter. Server license acceptance remains server-owned.
Qwen Image 2.1 (non-commercial Qwen Research License) edits from up to ten
ordered references, renders transparent PNG/WebP cut-outs, runs native 2K
presets, and ships BF16, INT8, FP8, GGUF and 6-step turbo tiers; its CUDA
fast path is 2.5x faster at 1024² and 6x at 2K than v0.32.
- **Images:** text-to-image, image editing, inpainting, ControlNet, LoRA,
identity photos, reference-image prompting, transparent backgrounds, prompt
expansion, and upscaling, as PNG, JPEG or still WebP.
- **Video and audio:** text/image-to-video, sequences, clip continuation,
lip dub, text-to-audio, and MP4 output with generated audio.
- **3D:** single-image and named multiview-to-mesh with Hunyuan3D 2.0 and 2.1,
automatic background removal, optional highlight and lighting removal, and