Meterless.ai
Run AI locally and own the whole workflow
73 followers
Run AI locally and own the whole workflow
73 followers
Most AI tools return an answer and discard the process. Meterless lets the real workflow run on your device—and makes it yours. Relay turns desktop work into reusable missions. Gaia keeps projects, memory, and context persistent. Swarms exposes the full agent graph. Replay, edit, switch models, and run the work again without starting over. Local-first, model-agnostic, and built so AI work compounds instead of disappearing.
DiffSense
What kind of hardware do you need to run this?
@conduit_design it runs in browser on any laptop or computer. For desktop control only need Windows
DiffSense
@samfrommeterless Aha. okay. so the harness sort of runs local and then you use remote AI models?
@conduit_design Each product lets you run with local or remote models but the main key unique part of Meterless both product store the actual work the AI does to be editable, scheduled or rerun with different models.
So for Relay for example you give it access to any window on your computer give it a prompt of what you want it to do. It will compile a mission for completing it. That mission lives on your computer so you can rerun it any time without models or with cheaper ones. You can refine it make changes or schedule to run anytime.
Gaia does the same kind of storing the throughput locally but it acts more as a replacement for something like Claude Cowork or Openworker.
That is the key is it builds a world model of what jobs missions and more you want to complete.
DiffSense
@samfrommeterless Love it! love the world concept. And the continuity perspective. I think it would be interesting to see hard evidence. Like benchmarks for tasks completed etc. Compared to other harnesses etc.
Hi Samuel, the part that stuck with me is that all this work no longer just vanishes when you close things. That really annoys me, so seeing someone treat it as worth keeping feels right, genuinely curious to see where this goes.
@robin_de_lacroix Thanks this is just the beginning. We are just getting started. MUCH MORE releases and tools to come. Our core bet is the industry is focused on the wrong thing. While everyone is focused on AIs outputs we are focued on its throughput.