Your own private computer
Every account comes with a secure cloud VM on a global edge network. It stores and manages your website cookies, access credentials, API keys, logs, traces, private knowledge and documents. Dedicated hardware soon
OMuse is an open source alternative to Meta’s Muse. Your cookies, credentials, API keys and documents stay on a machine that belongs to you.
Regular price is $29 a month once OMuse officially launches.
OMuse does what a personal assistant should do, and it keeps everything it touches on a computer you control.
Every account comes with a secure cloud VM on a global edge network. It stores and manages your website cookies, access credentials, API keys, logs, traces, private knowledge and documents. Dedicated hardware soon
OMuse works in the browser on your computer. You can see exactly what the agent sees on the screen at any time, and take over for the human steps such as logging in to a session.
Install any skill on your computer. A simple skills page lets you add, review and switch them on or off.
Create and edit what OMuse remembers and knows. Ask it to deep research any topic, and it puts the report into your own knowledge base.
Use the model and inference provider you like and trust. OpenAI inference is already included in the monthly subscription.
The full source code is on GitHub under the MIT license. You can deploy your own OMuse for free.
An independent guard service decides allow, deny or ask for every action the agent takes.
App passwords and tokens are encrypted inside Sentinel. The agent and the model never see them.
Emails, web pages and tool output are treated as untrusted. Suspicious content sends every write to you for approval.
An append-only, hash-chained log records every model call, tool call, decision and approval.
OMuse has the complete agenda and speaker profiles for GOSIM and Rust China Conf in Shenzhen. Ask it to search or plan any session, and use it to socialize with the speakers you want to reach out to.
It has current knowledge of Shenzhen’s life and tech scenes. Use it to plan your stay.
Subscribe during the introductory period and your price stays at $1 a month for as long as you keep your subscription.
Deploy the open source code on your own machine.
The price stays at $1 while you stay subscribed.
The price for new subscribers once OMuse officially launches.
One container holds every service, and all state lives in a single volume. Point it at any OpenAI-compatible model endpoint.
# build the image git clone https://github.com/Drlucaslu/locius.git && cd locius docker build -t omuse . # run it, with state in the omuse-data volume docker run -d --name omuse --restart unless-stopped \ -p 8080:8080 -v omuse-data:/omuse \ -e OMUSE_PASSWORD='choose-a-long-password' \ -e OMUSE_MODEL_URL=https://api.openai.com/v1 \ -e OMUSE_MODEL=gpt-4.1 \ -e OMUSE_MODEL_API_KEY=sk-... \ omuse
Open http://localhost:8080 and sign in as omuse. Put HTTPS in front before you expose the port to the internet.
The price stays at $1 for as long as you keep your subscription. Regular price is $29 a month at official launch.
Start for $1 a month