warm your relationships
Embers drafts and sends iMessages in your voice,
tuned to each person, so conversations with family, new friends, and
important leads never go cold when life gets busy.
🔥 Get early access & lock-in founders pricing · Coming Summer 2026
one bot,
five kinds of glow.
Timely replies in your voice, sent from your phone. Unique for each person in your life. Embers learns how you text each specific person. Mom doesn't sound like your top lead. Your brother doesn't sound like Alex from marketing. Embers handles the nuances gracefully.
Family
Parents, partner, siblings. The threads you want warm forever, even the weeks you're underwater.
Leads
Realtor leads. Recruiter candidates. SDR prospects. The pipeline that goes cold when you don't reply in 48 hours. Every prevented lead-death is real money.
Networking
The person you met at that conference. The friend of a friend from the party. The warm intro that fizzles. Embers keeps it alive until the next coffee.
The friend who sends 47 texts before noon
You love them. You can't always reply. Keep the friendship alive without burning yourself out.
Re-engagement
The investor you promised a quarterly update. The alumni you meant to ping. The ex-customer you liked. Embers sends the check-in in your voice, on a cadence you set — and the relationships compound.
Built for one relationship —
or five hundred.
No spam. Every thread is opt-in. Every draft uses the right voice. You stay in the driver's seat — Embers just makes sure the fire never goes out.
Per-thread voice NEW
You text your mom differently than your top lead. Embers knows. Onboard each contact with a tone sample — warm for mom, bold for your brother, professional for Alex. Every draft uses the right voice.
Catch-up NEW
Been slammed for two weeks? Ask Embers to catch you up on any thread before you jump back in. 3-sentence TL;DR of what's happened, what matters, what's still waiting.
Your Mac, your data
Reads your Messages database read-only, locally. Nothing stored on our servers beyond the moment a reply is drafted. Privacy-Hard mode runs inference on your Mac too.
Never initiates
Only replies to incoming messages you haven't answered. Never starts a conversation. Never double-sends. Never touches a thread you didn't add.
Menu-bar control
Pause any thread in one click. Review-before-send mode for the high-stakes ones. No terminal, no config files, no bash scripts. Ever.
Cadence you choose
Five minutes or five hours. Per-thread. Fast for leads, relaxed for family. Quiet hours supported — Embers sleeps when you sleep.
"The thread with my dad stays alive even the weeks I'm underwater. That's the whole point."
— THE REASON THIS EXISTS
Two ways to light it.
Self-Hosted — you run Embers on your own Mac, blue-bubble iMessage comes standard, we charge less. Cloud — we host everything, green-bubble SMS on any phone including Android, you pay a bit more to cover our SMS bill. Same features, same Catch-up, same per-thread voice. Every tier shows both prices. Pick what fits.
Try one thread. No card required.
- 1 thread · 20 replies/mo
- Self-Hosted or Cloud
- Per-thread voice + Catch-up
- "Sent via Embers" watermark
Personal — one or two important threads.
- 3 threads · 300 replies/mo
- Per-thread voice
- Catch-up included
The default — family, friends, a few leads.
- 10 threads · 1,500 replies/mo
- Tone samples + voice training
- Catch-up + weekly summaries
- Review-before-send
Solo pro — realtor, recruiter, coach.
- 50 threads · 5,000 replies/mo
- Bulk voice calibration
- Cadence rules per thread
- Priority support
Power user — heavy pipeline.
- 250 threads · 25,000 replies/mo
- HubSpot + Pipedrive sync
- Analytics dashboard
- Privacy-Hard mode
Teams — sales, recruiting.
- Unlimited · 5-seat min
- Shared voice training
- Salesforce + HubSpot
- SSO + team analytics
🔒 Waitlist signups lock founders pricing — your tier stays at this price forever.
Annual billing = ~2 months free at launch. Enterprise (SSO, SOC 2, on-prem Ollama): hello@tryembers.com.
Real questions.
Real answers.
We'd rather tell you what this isn't than oversell what it is.
When does Embers launch?
Coming Summer 2026. Spark ships first (Self-Hosted and Cloud), followed by Ember and Hearth within the first month. Bonfire and Forge land in phase two. Waitlist signups get early access + founders pricing locked at signup — your tier stays at that price forever, even when post-launch prices go up.
Self-Hosted vs Cloud — which one should I pick?
Self-Hosted runs Embers on your own Mac — you get blue-bubble iMessage out of the box and we charge less because you're running the bridge. Cloud is for when you don't want to run anything locally or you need SMS reach on Android recipients — we host the bridge and cover the SMS cost. Features are identical on both. Not sure? Pick "undecided" on the waitlist and we'll walk you through it at launch.
Is this going to pretend to be me? Isn't that weird?
It sends replies in your voice, from your Mac (Self-Hosted) or our bridge (Cloud), from your phone number. That's the point — the thread stays warm. Some people tell the person. Some don't. That's on you. Embers isn't trying to pretend to be you forever — it's trying to keep the thread alive until you show up in it again.
How does "per-thread voice" actually work?
When you add a thread, Embers asks how you text that specific person. Short and dry? Warm and exclamation-heavy? Professional? You describe it in plain English, and you can paste up to 10 of your own past messages for a tighter match. Every reply to that thread uses only that voice.
What is Catch-up?
When you've been heads-down for a week, open Embers' menu bar, tap "catch me up on Dad" — you get a 3-sentence summary of what's happened since you last engaged. What's the news? What's he waiting on? What's the vibe? It keeps YOU in the loop, not just the thread.
Where does my data go?
Self-Hosted: your iMessage history stays on your Mac. Embers reads it read-only. When a reply needs to be drafted, it sends the last handful of messages in that one thread to our proxy, which calls an LLM and returns a reply. Token counts stored, not message content. Privacy-Hard mode (Bonfire+) runs inference locally — nothing leaves at all. Cloud: same draft-time flow, bridge hosted by us, same no-retention policy on content.
Can I use this for work?
Yes — that's what Hearth, Bonfire, and Forge are for. Realtors run their lead pipeline. Recruiters warm candidates. SDRs keep sequences alive. CRM sync on Bonfire (HubSpot, Pipedrive) and Forge (Salesforce, HubSpot). Check local regulations on AI-assisted outreach disclosure — some states and industries have rules, and they're on you to follow.