Skip to content
simplexai-labsPublic

About

Linger is An Open-Source Personal Agent That Runs Fully on Your PC. We aim to break Big Tech’s grip on AI by bringing powerful local AI to everyday computers through efficient model–harness co-design, so everyone can run their own.

Topics

Resources

Contributing

Security policy

Stars

112 stars

Watchers

1 watching

Forks

Latest commit

 

History

163 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Linger

Linger is an open-source personal agent that runs locally on your computer.

Open-Source  ·  Local AI  ·  Consumer Macs & PCs

Documentation Powered by PolyStrata Powered by Strata

Follow on X YouTube Reddit

Linger: Open Source Personal Agent That Fully Runs on Your Local PC. Finds what you need across apps, remembers what matters, and gets things done. Open-Source, Local AI, Consumer Macs and PCs. Five pages play in turn. Local AI, powered by PolyStrata: a 125B parameter model on a 12 GB GPU, with the 55 GB 3-bit model file kept partly in GPU memory and the rest in RAM or on SSD, writing 45.5 tokens per second with Qwen3.8-Flash-Next on an RTX 3060 12 GB with 64 GB of RAM. Find and connect: stop searching app by app; Linger searches Files, Mail, Calendar, Slack and Docs and returns a project brief with its status, what is blocked and who can help. Remember and grow: a detail from a chat three weeks ago, Mom's birthday is the 14th and she loves orchids, is saved to memory and becomes a birthday plan with an orchid order to approve, right on time. Get it done: Linger keeps working while you are away, sorting the week's files by project, summarizing progress, drafting the recap email and preparing the weekly brief, ready for your review. Agent teams: one launch brief becomes a team.md that a research agent, a copy agent and a visuals agent follow, and they bring back a launch package for review

Powered by Efficient AI model–harness co-design, it brings powerful local AI to everyday computers. It understands your context across files and apps, remembers what matters, and proactively gets things done.

Your data and memories stay private and local.

Why local

Cloud chat assistant Linger
Where the model runs A remote server Your own GPU or Mac, on supported hardware
Where memory lives The provider's account On your device, in local mode
What it can reach What you paste in Files and apps you authorize
What you get back An answer A finished task, ready for review
Who can read the code The vendor Everyone. It is open source

Cool features

Run a 125B model on a 12 GB GPU, powered by PolyStrata: Qwen3.8-Flash-Next, whose 3-bit model file is 55 GB. GPU only, the baseline: the whole model has to sit on the card, only 12 GB fits and the rest has nowhere to go, so it needs 55 GB or more of VRAM, an H100 80 GB. With PolyStrata, ours: a router picks experts for each token. The busiest experts stay in the 12 GB of GPU memory, warm experts wait in 64 GB of system RAM and cold experts on the SSD. Tokens use experts on the GPU and in RAM, and an expert picked from the SSD is read into RAM while one RAM expert moves back to the SSD. Result: 45.5 tokens per second on an RTX 3060 12 GB with 64 GB of RAM, and 179 on an RTX 5090 32 GB

Feature What it does
Efficient AI model–harness co-design Brings a full local agent to supported consumer Macs and PCs.
Efficient MoE inference PolyStrata optimizes local inference so larger models can run with less GPU memory.
Privacy and data ownership In local mode, personal data and memories stay on your device. Connected apps exchange data with their services.
Cross-app context Understands work across authorized files and apps.
Wiki and memory Remembers important information, preferences, and decisions across conversations, in wiki pages every agent can search.
Proactive Another Me Watches your inbox, calendar and Slack, follows up and moves work forward, and messages you only when something needs you. Every "not now" teaches it.
Agent teams Brings in the agents a job needs and writes one team.md that they all follow.

Quick start

Option 1: the desktop app

For one person on one computer. Download the package for your platform from GitHub Releases. The app opens Linger, connects this computer and runs its agents. Quick Install on its setup screen sets up a Linger server on this computer too.

Platform Package
macOS, Apple silicon Linger-<version>-arm64.dmg
Windows x64 Linger-<version>-windows-x64-setup.exe, or -windows-x64-portable.zip to run without installing
Linux x64 Linger-<version>-linux-x64.tar.gz (glibc 2.35 or newer)

There is no package for Intel Macs; use Docker Compose there. Check each download against SHA256SUMS.txt from the same release.

Option 2: Docker Compose

For a server that several computers share, or to build from source. You need Docker with Compose and at least 5 GiB of memory for its virtual machine.

git clone https://github.com/simplexai-labs/linger.git
cd linger
cp .env.example .env
docker compose up --build --detach --wait

Create the owner account. As written, it signs in as owner@linger.local with the test password test1234; replace either one to use your own:

printf '%s\n' 'test1234' | docker compose exec -T teamhub \
  teamhub-account create-owner --email owner@linger.local --password-stdin

Change the test password in Settings › Security before other people can reach the server.

Open http://localhost:3000 and sign in. The app then shows the commands that connect your computer. See Install with Docker Compose to serve other computers on your network and to turn on voice.

Learn more

The full documentation lives at docs.heylinger.com.

 Set up What it covers
Desktop app Install the app, or build it from source.
Choose a runtime Claude Code, Codex or Pi, on local models or a subscription.
Self-hosting Run it with Docker Compose to serve several computers.
Troubleshooting Fixes for when something does not start.

Contributors

Linger contributors

New here? Read the contribution guide and pick a good first issue.

Community & Contact

Choose the channel that best fits your question:

Acknowledgements

Linger is built on these open-source projects:

  • PolyStrata: Runs local models, keeping the experts a model uses most on your GPU.
  • Strata: The inference engine PolyStrata is built on. It runs Qwen3.8-Flash-Next.
  • BlockNote, Yjs and Hocuspocus: Power the documents that people and agents edit together.
  • Hindsight: Where agent memory's fact extraction and search come from.
  • Cua: Lets agents use the apps on your desktop.
  • LiveKit: Carries the audio for voice and meetings.
  • River: Runs TeamHub's background jobs.
  • NOTICE: Lists every third-party component, what we adapted from it, and the license that applies.

Thank you to the open-source community for libraries this powerful and this simple: they let us focus on the product itself, and we hope our projects will in turn give everyone an easier-to-use knowledge base.

Feature Request

Have an idea? Open a feature request and describe the problem it would solve.

License

This repository is licensed under the Linger Open Source License, based on Apache 2.0 with additional conditions. NOTICE lists third-party code and the logos and images that this license does not cover.

Star History

Star History Chart

About

Linger is An Open-Source Personal Agent That Runs Fully on Your PC. We aim to break Big Tech’s grip on AI by bringing powerful local AI to everyday computers through efficient model–harness co-design, so everyone can run their own.

Topics

Resources

Contributing

Security policy

Stars

112 stars

Watchers

1 watching

Forks

Releases

Packages

Used by

Contributors

Languages