v0.1//Speech-to-text · macOS · Windows · Linux
Capture every
word.
Brassens écoute, transcrit et catalogue chaque mot que vous dictez — modèles à la pointe, multiplateforme, et toujours à portée d'un raccourci.
Coming soon · join the open beta
38 MB · .dmgApple Silicon · 12+v0.3.1 · open beta
02 // LIVE DEMO
Press to speak. Watch words land.
simulation · realtime preview
01 // Capture
● LIVE
Listening · tap to stop
or hitENTER
02 // Transcript
RECORDING · 00:00
00:01
00:04
————————————————
00:09
————————————————
00:14
————————————————
00:18
————————————————
00:23
————————————————
01 // Latency
~140ms
02 // Accuracy
98.4%
03 // Languages
99supported
04 // Footprint
42MB
Hosted models, no keys to manage//End-to-end encrypted//Local-first by default//Open beta · v0.1
03 // PROVIDERS
Pick the model that fits the moment.
included with pro
01 // PROVIDER
OpenAI
Whisper · GPT-4o Transcribe · Realtime
02 // PROVIDER
Deepgram
Nova-2 · Nova-3 · Flux
03 // PROVIDER
Mistral
Voxtral Mini · Voxtral Realtime
04 // PROVIDER
AssemblyAI
Universal · Multilingual