https://github.com/scottconverse/local-ai-control-center
Windows desktop control center for running Ollama, OpenHands, and Open WebUI locally without living in the terminal.
https://github.com/scottconverse/local-ai-control-center
electron local-ai ollama open-webui openhands windows
Last synced: about 1 month ago
JSON representation
Windows desktop control center for running Ollama, OpenHands, and Open WebUI locally without living in the terminal.
- Host: GitHub
- URL: https://github.com/scottconverse/local-ai-control-center
- Owner: scottconverse
- License: mit
- Created: 2026-05-14T05:22:15.000Z (2 months ago)
- Default Branch: main
- Last Pushed: 2026-06-11T20:33:30.000Z (about 1 month ago)
- Last Synced: 2026-06-11T22:14:47.144Z (about 1 month ago)
- Topics: electron, local-ai, ollama, open-webui, openhands, windows
- Language: TypeScript
- Homepage: https://scottconverse.github.io/local-ai-control-center/landing/
- Size: 1.34 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
Awesome Lists containing this project
README
# Local AI Control Center
**Published landing page:** [https://scottconverse.github.io/local-ai-control-center/landing/](https://scottconverse.github.io/local-ai-control-center/landing/)
**Latest Windows download:** [v0.5.5 release](https://github.com/scottconverse/local-ai-control-center/releases/tag/v0.5.5)
**User manual:** [USER_MANUAL.md](USER_MANUAL.md)
Local AI Control Center is a Windows desktop app for running Ollama, OpenHands, and Open WebUI locally without living in the terminal.
It is Windows-only for now and binds local service ports to `127.0.0.1` so OpenHands and Open WebUI are not intentionally exposed to your local network.
It gives you one GUI for:
- Running a first-run setup wizard.
- Reading the machine's CPU, RAM, disk, GPU, and VRAM.
- Installing Docker Desktop and Ollama from the app when possible.
- Pulling required Ollama models and Docker images.
- Streaming setup logs while long installs and downloads run.
- Remembering completed setup and opening to the Dashboard afterward.
- Starting and stopping OpenHands.
- Starting and stopping Open WebUI.
- Switching configured local models and resetting service data from the GUI.
- Editable labels and model-name recommendations for custom local model libraries.
- Checking Docker, Ollama, and service health.
- Opening the writable agent workspace.
- Opening the user manual.
- Embedding the OpenHands and Open WebUI interfaces in one place.
## Start The App
Download the latest portable Windows release, then open the app. On first launch, start with the **First Run** tab.
The first-run wizard checks whether the machine can comfortably run the stack. For the recommended OpenHands-ready local models, plan on:
- Windows 10/11.
- NVIDIA GPU with roughly 16 GB VRAM available.
- 32 GB system RAM.
- 80 GB free disk space.
- Internet access for first-time downloads.
From the source app folder, double-click the no-terminal launcher:
```text
Launch Local AI Control Center.vbs
```
If you want to see startup logs, double-click:
```text
Start Local AI Control Center.cmd
```
First-time setup:
```text
Install Local AI Control Center.cmd
```
That source setup helper installs Node dependencies and builds the app. End users who download the release EXE should use the in-app **First Run** wizard instead.
## User Manual
Full walkthrough:
```text
USER_MANUAL.md
```
## Landing Page
Published product page:
[https://scottconverse.github.io/local-ai-control-center/landing/](https://scottconverse.github.io/local-ai-control-center/landing/)
The source for that page lives at:
```text
landing\index.html
```
The page is intentionally honest about the current product state: useful local Windows control center today, still being hardened for wider distribution.
## Local Services
OpenHands runs at `http://localhost:3000`.
Open WebUI runs at `http://localhost:8080`.
Ollama runs at `http://127.0.0.1:11434`.
Docker is resolved from `DOCKER_EXE`, `DOCKER_PATH`, the system `PATH`, or the standard Docker Desktop install location.
The setup wizard can install Docker Desktop and Ollama through `winget` when Windows has it available. If `winget` is missing, the app opens the official download page so the user can install the missing tool without using a terminal.
## OpenHands
OpenHands is the local coding/agent UI. It can work in a Docker sandbox, read and write project files, and run commands.
Writable local workspace:
```text
.\agent-workspace
```
Inside OpenHands, that folder appears as:
```text
/workspace
```
Start it:
```powershell
.\start-openhands.ps1
```
Then open:
```text
http://localhost:3000
```
Recommended first settings for agent/file work:
```text
LLM Provider: OpenAI-compatible
Model: openai/qwen2.5-coder:14b
Base URL: http://host.docker.internal:11434/v1
API Key: local-key
```
Gemma is still available for fast local chat:
```text
gemma4-26b-8k
```
## Open WebUI
Open WebUI is a friendly local chat interface for Ollama.
Start it:
```powershell
.\start-openwebui.ps1
```
Then open:
```text
http://localhost:8080
```
## Development
Developer setup, packaging, and maintainer commands live in:
```text
DEVELOPER.md
```
Install dependencies:
```powershell
npm install
```
Run the desktop app in development mode:
```powershell
npm run dev
```
Run the test runner:
```powershell
npm run test:runner
```
Run build/unit checks without requiring local services:
```powershell
npm run test:unit
```
Optional local LLM smoke test:
```powershell
npm run test:llm
```
## Local Runner
Use the local runner before publishing or packaging changes.
Use the local runner first:
```text
Run Local Test Runner.vbs
```
See:
```text
RUNNER.md
```