Nymbus
My Apps

iOS

Nymbus

Your models, from anywhere. Chat, agents and orchestration for LM Studio and the cloud.

Coming soon iOS · Free

What it does

Overview

The model is on your desk, not your phone

A 35B model will never run on an iPhone. But the Mac on your desk is already running one.

Nymbus is a client, not a host. Point it at LM Studio on your own computer and chat with the models you've downloaded, from the couch or across town over Tailscale. Want a bigger model for one job? Add an OpenRouter, Requesty or Cerebras key and switch between local and cloud whenever you like.

No accounts, no analytics, no tracking. Nymbus only talks to the servers and providers you set up.

A chat with a local Qwen model planning three days in Lisbon
An agent run with web search, a page fetch, a file write and a plan checklist
The tide pool guide the agent built, open in the live preview
Ask, and the agent searches, plans, writes the files and checks its own work. The page it builds opens right in the app.
The LM Studio model library with downloads and loaded models
An orchestrator run with builder, designer and reviewer workers
The usage dashboard with tokens, speed and per-model tables
Manage your model library, let one model run a team of workers, and see where every token goes.

Status

On its way to the App Store

Nymbus is getting ready for its first App Store release, iPhone only. You'll need LM Studio on a Mac or PC, or an API key from a supported cloud provider. The case study covers how it's built.

More from My Apps

macOS

Nodebook

AI chat on an infinite canvas. Branch a conversation instead of losing it.

Web

Redline

Mark up a screenshot and hand the critique to a coding agent.

Web

Backdraft

Import your codebase, edit it on a visual canvas, and every change writes back to your real source.