At a Glance
- Todoist Ramble converts natural speech into organized tasks in real time
- Built on Google Gemini 2.5 Flash Live, it supports 38 languages across iOS, Android, desktop, and web
- Beta testers completed 290,000 sessions in three weeks, pushing task-creation success from 40% to 62%
- Why it matters: Users can capture to-dos hands-free without typing, narrowing the gap between thought and action
Todoist’s parent company Doist is launching Todoist Ramble, an AI-powered “voice-to-tasks” feature that lets users speak naturally and watch their words become structured tasks, deadlines, priorities, and assignees without typing a single character.
How Ramble Works
Tap the Ramble icon inside the Todoist app and start talking. The moment you say “Actually, make that Thursday” or “that’s all,” the interface updates live, reshaping tasks as you refine them. Ramble streams audio to Google’s Gemini 2.5 Flash Live model on Vertex AI, transcribing speech and extracting actionable details on the fly.
Key capabilities:
- Accepts meandering, unstructured speech
- Captures deadlines, priorities, duration, assignees
- Edits tasks in real time as you speak
- Audio is not stored or used for model training
- SOC2 Type II certified, according to Doist
From Beta to Global Rollout
Ramble first reached Todoist Experimental users-about 150,000 active testers. Within the first three weeks:
- 76,000 testers completed roughly 290,000 sessions on desktop and mobile
- Task-creation success rose from 40% in October to 62% by December
- New entry-level users upgraded at roughly five times the pre-Ramble rate
Public beta testing began in November. Today the feature exits beta and ships to every Todoist user across iOS, Android, desktop, and web, supporting 38 languages.
Access and Limits
| Plan Type | Ramble Sessions |
|---|---|
| Beginner | Limited per month |
| Pro | Unlimited |
| Business | Unlimited |
Quick-access options:
- iOS Home or Lock Screen shortcut
- Android app shortcuts, widgets, quick-settings tiles
The Bigger Picture
Ramble joins Todoist Assist, Doist’s umbrella of AI features. The launch parallels hardware trends where wearables transcribe speech into notes or to-dos. Amazon’s Bee wearable, Plaud’s meeting recorder, Sandbar’s Stream ring, and Pebble’s upcoming Index 01 ring all pursue the same goal-friction-free capture-but Ramble achieves it inside an app millions already use daily.
Key Takeaways
- No new hardware required; works inside existing Todoist apps
- Real-time understanding keeps pace with natural speech
- Early data shows faster upgrades and higher task-creation accuracy
- Multi-platform, multilingual availability positions Ramble as the widest-reaching voice-to-task tool released so far
