QuietQueue.
A LITTLE LESS, A LITTLE BETTER

Less juggling.
More done.

Give your scattered tasks a quieter rhythm.
A small plan, similar tasks together, room to breathe.

02 / FIND YOUR RHYTHM

Your next 30 minutes

Rule-based
01

Empty your head

Messy lists are welcome here.

02

Stay in one headspace

Related tasks, one gentle batch.

03

Leave the rest for later

A parking lot, not a guilt trip.

A QUIETER WAY TO WORK

Less switching.
More breathing room.

QuietQueue groups related tasks into a few small batches and fits a selection inside your time budget. The rest stays in a parking lot. Time estimates are planning aids, not predictions.

Useful without AI

The default planner uses transparent task categories and works without external requests. Your input and checkoffs are saved only in this browser when storage is available.

Open weights. Local inference.

The optional AI mode uses all-MiniLM-L6-v2 embeddings through Transformers.js. It downloads model and runtime files from external hosts, then compares task meanings in your browser. Your task text is never sent to a model server. If loading fails, the rule-based planner still works.

Honest, small numbers

“Switches avoided” compares the selected tasks in their original order with their batched order. It is a count of category changes, not measured time saved. Checking tasks off does not change the estimate.

No sign-in, API key, analytics, or productivity score. Source code is included with this project. “Clear local data” removes saved tasks and checkoffs, not model files cached by your browser.