Voice to Text for Roam Research
Your networked thoughts don't wait for you to finish typing. A connection sparks while you're reading, walking, or mid-conversation. Blurt lets you capture those insights directly into Roam Research before the thread unravels. Hold a button, speak your thought, release. Text appears in your block, ready for linking and referencing. No friction between insight and capture. Your graph grows faster when you can think out loud.
The Typing Problem
Daily notes become another inbox you avoid
You opened Roam with the best intentions. Today's daily note sits empty. You have thoughts worth capturing, but typing them out feels like work. So you tell yourself you'll catch up later. Tomorrow's daily note will also be empty. The habit never forms because the activation energy is too high. Your best thinking happens away from the keyboard, and by the time you sit down, the thoughts have scattered.
Block-based writing should feel atomic, not tedious
Roam's power is in the blocks. One thought, one block. But when you're on a roll, stopping to hit Enter between thoughts breaks your flow. You end up with massive paragraphs that need splitting later. Or worse, you write less because chunking feels like extra work. The outliner structure that should liberate your thinking becomes a formatting chore.
Bi-directional links require typing the same context twice
You want to reference [[Project Alpha]] in your daily note. But just dropping the link isn't enough. You need to explain why it's relevant today. What you noticed. How it connects. So you type out the context. Then later, when you visit Project Alpha, you'll type similar context again. The linking is instant; the meaning-making takes forever.
Research notes pile up faster than you can process them
You're reading papers, articles, books. Ideas are firing. But capturing them in Roam means stopping, switching windows, typing, formatting. By the time you've documented one insight, three more have evaporated. Your reading sessions produce half the notes they should because capture can't keep pace with comprehension.
Journaling in Roam feels like documentation, not reflection
Morning pages should be stream of consciousness. But typing forces you to think about spelling, backspacing errors, fixing autocorrect. You're editing while you should be exploring. The journal entries that should reveal your thinking become sanitized summaries. You write what's easy to type, not what's true to think.
How It Works
Blurt works directly in Roam Research. In your daily notes, any page, any block. No extensions needed.
Hold your hotkey
Press your chosen shortcut. A small indicator shows Blurt is listening.
Think out loud
Speak your thought, reference, or connection. Blurt handles punctuation automatically.
Release and continue
Text appears in your current block. Add your brackets and links, or keep capturing more thoughts.
Real Scenarios
Morning journaling in daily notes
Your morning coffee is ready. Today's daily note is open. Instead of staring at a blinking cursor, hold the button and let your thoughts flow. 'Woke up thinking about that conversation with Marcus. Something about how he frames decisions as experiments not commitments. Connects to my note on reversible versus irreversible choices. Need to revisit that framework for the product launch decision.' Thirty seconds of speaking captures what would take five minutes to type.
Capturing ideas while reading
You're deep in an article and a connection fires. Don't lose it. Hold the button: 'This argument about network effects mirrors what I read in that economics paper. The key difference is they're talking about demand-side rather than supply-side. Should link this to my note on platform strategy.' The insight is captured while it's hot. Keep reading without breaking flow.
Building block references quickly
You need to reference something from last week's daily note in today's context. Hold and speak: 'Following up on the client feedback from Thursday. They specifically mentioned the onboarding flow was confusing. This validates what Sarah said in our design review.' The context is captured. Add your block reference after, with the meaning already documented.
Processing meeting notes into networked thoughts
After a meeting, your notes are linear. They need to become networked. Go through each point and speak the connections: 'This action item about the API documentation connects to the developer experience initiative. The timeline concern links to our Q2 planning note. Assign to Marcus, deadline is the fifteenth.' Raw notes become linked thoughts in minutes, not hours.
Expanding linked references with context
You've linked [[Decision Frameworks]] but the link is just a pointer. Hold and speak: 'Using the decision framework here because this choice is high stakes and reversible. The key criteria are speed to market and team capacity. Leaning toward option B because it optimizes for learning.' Now the link has meaning. Future you will understand why this reference mattered.
Daily review and reflection
End of day. Time to review what happened. Hold the button and process out loud: 'Good day overall. Made progress on the proposal but got stuck on the pricing section. Tomorrow I need to finish that before the eleven AM call. Feeling better about the project direction after talking to the team.' Your daily review becomes a conversation with yourself, not a typing exercise.
Research synthesis across sources
You've read three papers on the same topic. Time to synthesize. Hold and speak: 'All three authors agree on the importance of iteration speed, but they differ on how to measure it. Paper one uses cycle time, paper two uses deployment frequency, paper three uses customer feedback loops. My synthesis is that the metric should match the feedback you're optimizing for.' Synthesis captured while the comparisons are fresh in your mind.
Why Roam Research users choose Blurt over built-in dictation
| Blurt | macOS Dictation | |
|---|---|---|
| Activation | Single hotkey, instant start | Click microphone or double-tap Fn key |
| Speed | Text appears in under 500ms | 2-3 second delay before transcription |
| Reliability | Consistent accuracy every time | Often fails silently or mishears |
| Natural speech | Handles thinking-out-loud style well | Works best with formal dictation style |
Frequently Asked Questions
Start Typing Faster Today
Free to try — no credit card required
Download Blurt