mere.run - Local-first CLI for AI workflows without Python
Show HN: Local text, image, video, music and 3D from one CLI, no Python
mere.run is a powerful local-first inference runtime designed for Apple Silicon and headless Linux, enabling users to execute text, image, video, music, and 3D workflows from a single command-line interface. Built entirely in Swift without Python dependencies, it offers an OpenAI-compatible API server and a unified model management system. Whether you need to generate images, transcribe speech, create 3D assets, or run local agents, mere.run handles it all on your machine with optional macOS Studio integration. It supports distributed computing via Relay and Nodes, making it a versatile tool for developers and creators seeking privacy and performance.
One real terminal session on an M4 Max — chat, image, grounding, image→3D, SFX, music, speech round-trip, video, API serving, and a typed workflow graph. Everything local.
- sawfwair
some real numbers on my m4 max -
image gen (zimage-nano), 1024x1024: 58s
image -> textured mesh (trellis.2): 2m 49s
sfx generate (5s clip): 3.6s
music generate (8s, ace-step): 15s
speech synth: 13s | transcribed back: 2.2s
video gen w/ audio(ltx unified-av) 4s,768x512: 2m 48s
text chat (laguna xs2.1): 102 tok/s - https://mlx.fast leaderboard
- ks2048
FYI, the title makes no sense in isolation.