Create anything locally

Generate images, video, worlds, 3D, music, speech, and sound. Train LoRAs. Inspect scenes. Use the Mac app, Linux CLI, or local API.

Downloads Linux QuickStart GitHub source MIT licensed · release assets from GitHub
macOS app Linux x86_64 Linux aarch64 CUDA source arm64 runtime builds from source OpenAI-compatible API Relay nodes Official plugins
The gallery

Made with Mere

Each tile shows output from one MacBook Pro. Hold the pointer over a tile, focus it, or tap it to view the command.

Explore the full showcase →
Create · Video and sound
$ mere.run video generate --variant unified-av --model video-ltx23-av-mlx
LTX 2.3 · image → video loop · synced generated audio · seed 2121 · M4 Max · deep dive →
Create · Image Locally generated portrait of a woman in window light.
$ mere.run image generate --model image-klein-nano
Klein · portrait · seed 1234
Create · Music
Waveform of a locally generated music track.
▶Hear it
$ mere.run music generate --model music-acestep
ACE-Step v1.5 turbo · 35s · 8 steps · seed 2121
Reconstruct · 3D
$ mere.run vision image-to-3d-trellis2 fox-cutout.png
TRELLIS.2 · 2.2M-triangle textured mesh · turntable rendered offline · deep dive →
Analyze · Geometry Recovered depth map of the plate. The original generated plate.
platedepth
$ mere.run vision geometry plate.png -o ./geo
MoGe-2 · metric depth · drag to compare · deep dive →
Analyze · Vision Generated diner scene with Falcon Perception detection and segmentation overlays.
$ mere.run vision ground "waitress; jukebox; neon sign" diner.png
Falcon Perception · SAM 3.1 · detect + segment · deep dive →
Capabilities

One local-first runtime

Create, understand, train, reconstruct, automate, and serve from one scriptable local runtime. Every capability runs on your machine.

Create
Images, video, music, and voice

Generate and edit images with Klein, ZImage, and Krea 2. Create video with synchronized audio on LTX 2.5, navigate persistent worlds, sing lyrics with MiniMax Music 3, and add foley or speech.

mere.run video generate --variant unified-av --model video-ltx23-av-mlx See it run →
Understand
Vision, language, and retrieval

Ground, segment, track, caption, and read text with Falcon Perception, SAM 3.1, and LightOn OCR. Reason across modalities, embed for local RAG, and anonymize PII before anything leaves disk.

mere.run vision ground scene.png --query "waitress" See it run →
Train
Adapters and evaluation

Train Klein and Krea 2 image adapters or text adapters from SFT JSONL. Preflight checks the plan first, and evaluation runs publish reproducible result receipts.

mere.run image train-lora --data ./dataset --output ./style.safetensors --preflight --json See it run →
Reconstruct
3D, geometry, and Earth observation

Turn one image into sealed PBR meshes with TRELLIS.2, recover metric depth, normals, pose, and flow with MoGe-2, and encode Sentinel observations with TESSERA and OlmoEarth.

mere.run vision image-to-3d-trellis2 object.png -o ./mesh See it run →
Automate
Graphs, executors, and plugins

Author immutable portable graphs, then run the same plan locally, over SSH, or through Relay. Inspect runs, fetch artifacts, and extend everything with official plugins.

mere.run graph run workflow.json --inputs-json inputs.json --run-dir ./runs/local See it run →
Serve
Local API and apps

Serve OpenAI-compatible chat, embeddings, images, speech, and transcription on your own machine, connect Open WebUI, and manage models with memory-aware recommendations.

mere.run open-webui quickstart --pull API details →
mere.

Commands are grouped by modality and designed for scripts.

~/projects · zsh
$ mere.run model pull image-3d-trellis2-4b --preflight --json download size, disk, and license reported before any bytes move
$ mere.run vision image-to-3d-trellis2 ./object.png --seed 43 -o ./mesh sealed PBR mesh · GLB + OBJ + PLY + PBR voxels + run manifest
$ mere.run image train-lora --data ./dataset --output ./style.safetensors --preflight --json typed plan first; drop --preflight to train with loss CSV and checkpoints
$ mere.run video cosmos3 "navigate the room" --mode text-to-video --output ./world.mp4 Cosmos3 Edge generation and action dynamics
$ mere.run music transcribe ./mix.wav --format midi -o ./mix.mid MuScriptor instrument-separated MIDI from a full mix
$ mere.run graph run workflow.json --inputs-json inputs.json --run-dir ./runs/local immutable graph → portable execution → durable artifacts
Agent workflows

Plan, run,
and inspect

--preflight --json reports downloads, memory fit, and planned work before loading a model. Run immutable graphs locally, over SSH, or through Relay. Each approved run stores artifacts, diagnostics, and next actions on disk.

  • Preflight for training, video, SFX, tracking, serving, and model pulls
  • Reusable JSON run plans
  • Searchable run directories with manifests, checksums, and provenance
  • Model recommendations based on the machine
  • Resident API models with admission control and continuous batching
  • Portable graphs with watch, fetch, resume, and material provenance
agent session · preflight → approve → execute → inspect
$ mere.run image train-lora --data ./dataset -o ./style.safetensors --preflight --json "status": "ok" · "24 usable pair(s), ready to train." · plan typed out, nothing loaded
> plan approved
$ mere.run image train-lora --data ./dataset -o ./style.safetensors 144 layers injected · 250/250 steps · adapter + 4 checkpoints + loss CSV saved
$ mere.run run inspect ./run --json "status": "ok" · 41 artifacts · diagnostics + next actions
Open-source tools

Add machines
and workflows

mere.run provides local inference. Relay pools machines, Graph Studio creates portable visual workflows, and official plugins add production integrations. Each project is MIT-licensed and has public source.

01 Relay + Nodes

Pool the machines you already have

Install mere.run node on a Mac or Linux box. It reports its models and availability to your account, then accepts work through an outbound connection.

  • One private compute pool per mere.world account
  • Balanced, fastest, or power-efficient scheduling
  • Node downloads for macOS, Linux x86_64, and Linux arm64
  • Open source under the MIT License
install node → approve machine → send work
Open Relay Get a node GitHub source
02 Graph Studio

Compose on a canvas and run on your GPUs

Author typed image, video, audio, and text workflows visually, then execute the same portable graph on your workstation, over SSH, or through Relay.

  • Easy and Pro modes over one graph document
  • Executor-aware preflight, runs, and artifact inspection
  • Hosted Studio plus an offline desktop app
  • Open source under the MIT License
author → preflight → run on your GPUs
Open Graph Studio Desktop app GitHub source
03 Official plugins

Install the workflow you need

Plugins handle plans, artifacts, and cleanup around mere.run inference.

  • VFX, realtime music, ShotGrid, redaction, automation, and GPU training
  • Thirteen official plugins in a machine-readable catalog
  • Durable manifests, resumable runs, and explicit cleanup
  • Open source under the MIT License
$ mere.run plugin install mere-vfx-tools
Browse plugins Docs GitHub source

Runs on your hardware

No hosted runtime or per-call bill.

Data stays on disk

The runtime does not upload prompts or outputs. This public site records privacy-minimized product analytics.

MIT licensed

Source on GitHub. Built in Prince Edward Island.