Quartz - AI email client built for focus. Runs locally on your Mac

Quartz turns Gmail into a focused inbox. It sorts every message by importance, and learns what matters to you over time. When you reply, it drafts in your own voice. And the AI runs entirely on your own Mac, so your mail stays end-to-end encrypted and never shared with AI providers.

Add a comment

Replies

Best

Very excited to try this out! Would that send a push notification when an email is very important so I can respond asap?

, indeed we send notifications for emails that get into important category.

One thing I think we probably should showcase more is that you can change categories of emails manually and this will be remembered (locally, data is visible in settings for review) and would influence categorization of the following emails. If an important email slipped into non-important category, you can fix it manually and next time similar email would go to important and you'll see a notification.

 thank you! Yes, that's the whole idea. You won't see push notifications unless it's important. And "important" is defined by you, Quartz will get more accurate over time as it learns more about you!

Really interesting local first approach. Just curious, which model(s) are you running locally?
Also, can I define any custom categories? Like emails about credit card transactions or UPI payments?

 good questions!
We're running Gemma 4 E4B on-device. Custom categories aren't in yet, right now it's the standard five importance levels, but they're on our roadmap!

   while we don't have custom categories, it's possible to steer Quartz to put CC transactions and similar emails to particular category. You may describe desired behavior in AI Settings or you can manually define a category for such emails and Quartz will learn to use the category for similar emails in the future.

   - Cool, I'll try it out this week. Currently I add custom filters in gmail which is a manual process and doesn't adapt automatically.

It looks like there is a chance I'll finally start reading my emails :D

, that's cool to hear ❤️. Let us know if you have any feedback ;)

Our changelog for today so far

Congrats on the launch! This tool would definitely make my life easier. Does it support adding multiple emails/domains, to tackle everything at once?

, thank you ❤️

It supports adding multiple Gmail accounts. I personally have 2 Google Workspaces accounts and one personal Gmail account added.

Been dreaming about this for a while but for comms more broadly: telegram, whatsapp, etc. I think Eric (former YC partner) tried this. Good luck!

 Thank you! Email felt like the most painful place to start, but the long-term vision is much broader. I want Quartz to become a personalized attention layer that understands what’s important to you across email, messaging and eventually every channel competing for your attention

Focus and privacy in the same product is rare for an email client. Does it handle multiple accounts in one inbox?

, yes, it does. Unfortunately, only Gmail for now. I use multiple accounts myself.

Appreciate not having to add yet another service to the chain just to get the AI-powered goodies in the email client. Local LLM is the way. Congrats on the launch! 🚀

 thank you ❤️

Great project! You mentioned how cold emails create clutter and this tools helps dealing with that.

Do you think we can expand this into agent2agent communication? Like both sides essentially have AI messaging each other and if things make sense, the issue gets escalated to humans.

One of the reasons we treat cold emails as spam is because there are so many of them. However, if there is a proper processing layer, I actually do wish people to reach out to me and let me know how they can bring value.

 I think eventually we'll get to agent2agent communication: my personal AI agent will be speaking to yours and both of us will only get involved if there's something of common interest. My big vision for Quartz is building personal attention layer across all communication channels (email, WhatsApp, Telegram, etc), email is just the most painful channel as of now (no regards to privacy)

local AI for email is the right privacy architecture and the right differentiator against every other AI email client. the trust question for email tools is specifically about whether your correspondence is being used to train models or stored on someone else's servers. running on your own Mac removes that question entirely rather than asking users to trust a privacy policy

 to be honest, local-first approach also has some downsides: inference is slower and models aren't that powerful. But we think it's a fair tradeoff for email client.

 that's a fair tradeoff to name explicitly, and it's probably the right one for email specifically since most of what an email client does, sorting and drafting, isn't latency-critical the way something like a coding assistant would be. the model capability gap matters more for drafting quality than for sorting accuracy, I'd guess. is that roughly where you're seeing the tradeoff bite hardest too?