Doist, the company behind the popular task management app Todoist, has launched a new AI powered feature called Todoist Ramble, allowing users to create tasks simply by speaking. Designed to capture natural, unstructured speech, the feature turns spoken thoughts into organised to do lists, automatically identifying details such as deadlines, priorities, duration, and assignees as users talk.
Ramble is aimed at people who often think of tasks while on the move but find typing inconvenient. By tapping the Ramble icon, users can speak freely while the app updates tasks in real time, even responding to corrections like changing dates or priorities mid sentence. The feature is built on large language models under Todoist Assist, using Google’s Gemini 2.5 Flash Live model to transcribe speech and extract relevant task information without storing audio or using it for AI training.
The launch places Todoist alongside a growing wave of AI driven tools that convert speech into notes and reminders, a trend seen in emerging wearables and smart devices. However, Doist says early testing shows strong engagement within the app itself. During beta testing with around 150,000 experimental users, more than 76,000 testers completed nearly 290,000 Ramble sessions in just three weeks, while task creation success rates rose significantly toward the end of the testing period.
Todoist Ramble is now available to users across iOS, Android, desktop, and web platforms, with support for 38 languages. While beginner plan users have limited monthly sessions, Pro and Business subscribers can access the feature without restrictions. According to Doist, the voice driven experience has already boosted user upgrades, reinforcing the company’s belief that simpler, more natural interactions can meaningfully improve everyday productivity.