You can speak about 150 words a minute. You type maybe 40. Voice is the fastest way to steer a coding agent - but only if the transcription actually knows what kubectl and PostgreSQL are. That is where DevThrottle is different.
Because your hands and your screen are a bottleneck, and your agents move faster than that.
Dictation runs at roughly 150 words per minute against 40 for a good typist. A long instruction you would never bother typing takes ten seconds to say.
On a walk, in the car, making coffee. When an agent stops and asks a question, you answer out loud and it keeps shipping - no keyboard required.
Thinking out loud keeps you in the work. You describe what you want in plain language instead of stopping to carefully type a prompt.
Everyday dictation is great at everyday words. Developer talk is its own language - library names, CLI tools, your own project names. Give transcription a dictionary of those terms and the same sentence comes through exactly as you meant it.
When your instructions are precise, the words have to be precise too. The dictionary is what turns voice from a neat demo into something you reach for all day.
Teach it your words once: project names, library names, the jargon your team actually uses. It remembers, and that same dictionary follows you to every device you talk to your agents from.
Desktop · Cockpit · Phone
Add a term at your desk and your phone knows it on the next walk. No re-teaching, no per-device setup, no drift between machines.
Working in the Director app at your desk? Hold to talk and your words land in the right agent - with your project names spelled right every time.
Steering a dozen machines from the browser? The same dictionary rides along, so a term you taught at your desk is understood from the control room too.
Out for a walk and an agent needs an answer? Talk back from the mobile app. Same voice, same dictionary - no phone keyboard, no autocorrect fighting your jargon.
The more you use it, the better it gets. Every term you add makes voice sharper across every machine you own.
The custom dictionary and voice mode are part of the product. The only question is whose AI key does the transcription.
Bring your own AI key for voice.
We handle the key. Zero setup.
Join early access and be first to teach it your words.
Get Early Access - Free