CueInbox reads all your notifications, triages them by priority, and drafts smart replies — powered by on-device AI. Nothing ever leaves your phone.
Telegram pings. Slack threads. Gmail follow-ups. WhatsApp groups. LinkedIn messages. Every app fights for your attention, but only a few messages actually need you right now. There's no single place that answers: “What actually needs my attention?”
CueInbox captures notifications from all your messaging and productivity apps, runs them through AI that understands urgency, context, and intent — and presents a single priority queue.
AI classifies every notification into Now, Today, Delegate, or FYI. VIP contacts always surface. Noise gets filtered.
3 contextual reply options per message using the last 5 messages of conversation. Concise, warm, or direct. Tap to send.
Snooze any thread with AI-suggested timing. Get reminded when it matters, not when it arrives.
Every notification passes through a two-stage pipeline — instant heuristics first, then on-device LLM for hard cases. Chain-of-thought reasoning activates when confidence drops below 70%.
Text + image from any app
Instant, always works offline
Gemma 4 E2B on-device GPU
Now · Today · Delegate · FYI
Thread context, last 5 msgs
3 options per thread
Chain-of-thought reasoning activates automatically when confidence < 0.7 — the model explains its reasoning before scoring. Vision analysis runs on notification images and photo previews for richer context.
20 apps. One queue. Just grant notification access — no logins, no API keys.
“Finally an app that just works across all my messaging apps. The AI triage is surprisingly accurate — it correctly prioritized a critical work Slack message over a meme from a friend.”
“The privacy angle sold me immediately. Everything runs on-device — no API keys, no subscription, no data leaving my phone. This is how AI apps should be built.”
“I was drowning in 200+ notifications a day. After a week with CueInbox I only read what the AI surfaces as “Now”. Everything else waits. Game changer.”
Gemma 4 E2B — Google's latest language model — runs directly on your phone's GPU. 2.4 GB downloads once, then everything works offline.
| Todoist | Notion | Superhuman | Spark Mail | SaneBox | CueInbox | |
|---|---|---|---|---|---|---|
| Price | $7/mo | $10-20/mo | $33/mo | $8-20/mo | $7-36/mo | Free |
| Apps covered | Tasks only | Gmail only | Email only | Email only | Email only | 20 apps |
| AI triage | No | Business only | Auto Labels | Basic | Learns senders | Hybrid LLM |
| Draft replies | No | No | Auto Drafts | AI drafts | No | 3 per msg |
| Privacy | Cloud | Cloud | Cloud | Cloud | Cloud | On-device |
| Works offline | Partial | No | No | Queues | No | Full |
| Cross-app | No | No | No | No | No | Yes |
Download the APK and grant notification access. No account needed.
The Gemma 4 E2B model downloads once (~2.4 GB over Wi-Fi). Takes a few minutes.
Every notification gets triaged automatically. Open the app to see your priority queue.
CueInbox ships fast. Here's what's landed recently.
Download CueInbox and let AI handle the triage.
Have questions, feedback, or found a bug? We read every email.
info@cuelinbox.com