blurt/README.md

Blurt

Tray-based push-to-talk voice-to-text for Linux and Windows. Toggle recording with a hotkey, speak, toggle off — Blurt transcribes the clip and copies the text to your clipboard.

If a transcript starts with to my agent, the rest is dispatched as an agent task instead of copied.

Setup

Requires uv and Python 3.11+.

uv sync

API key

Blurt reads the key for its current model's provider in this order: environment variable → OS keyring → local file.

Provider	Env var	Keyring (service / user)	File
OpenAI	OPENAI_API_KEY	openai-api-key / felixm	api_key.txt
OpenRouter	OPENROUTER_API_KEY	openrouter-api-key / felixm	openrouter_key.txt

Key files and runtime config stay local (gitignored).

Running

Linux — uv run blurt starts the tray icon. Bind a hotkey to ./blurt-toggle, which signals the running instance to start/stop recording.

Windows — launch blurt.vbs (runs Blurt with no console window). The global hotkey is F20.

Configuration

Set from the tray menu, or via files / env vars:

• Model — model.txt / BLURT_TRANSCRIPTION_MODEL. For lowest latency use openai/whisper-large-v3-turbo (OpenRouter, served by Groq).
• Language — language.txt / BLURT_TRANSCRIPTION_LANGUAGE (en, de).
• Input device — tray menu (persisted by name in device.txt / BLURT_INPUT_DEVICE). Pick an explicit mic: PortAudio device indices shift when USB gear comes and goes, and the OS default source can change under you. The menu hides playback endpoints and raw ALSA hw: devices that the sound server already owns; BLURT_INPUT_DEVICE can still name one directly. Devices that refuse 16 kHz (raw ALSA and PipeWire's JACK endpoints do) are captured at their native rate and resampled.

Every clip is conditioned before upload — the stream-open click is trimmed off the front (only as much of it as is actually there, up to 150 ms), DC bias is filtered out, and the level is normalised towards -20 dBFS. Near-silent clips are refused instead of being sent, since transcription models answer silence with confident hallucinations.

Every clip is written to spool/ (gitignored) before it is uploaded, and deleted again as soon as its text reaches the clipboard. If transcription fails — timeout, connection error, server error, empty response — the WAV stays there and the notification says so; Retry failed transcriptions (N) in the tray menu re-uploads the spooled clips oldest first, so a network hiccup never costs you the words. A rejected request (any 4xx other than 408 and 429) is the exception: the same audio can only be rejected again, so it is not kept, and the notification says that too. Retried transcripts are copied to the clipboard together, oldest first, once the whole run is done — and never dispatched to an agent, since a spooled clip may be one part of a longer recording. The upload timeout scales with the clip: 30 s plus 30 s per minute of audio, capped at 5 minutes. Clips longer than 5 minutes are split into chunks of about 4 minutes — each cut lands in a pause, not mid-word — and the chunks are uploaded in parallel and stitched back together in order. The tray icon fills clockwise as each chunk comes back, so you can see how far along the clip is. If only some chunks come back, the transcript is copied with […] where the gaps are, and just the failed chunks are kept in the spool as …-partNN.wav for the retry.

Nothing else is cut: every second of the recording is uploaded. If a transcript comes back missing words, check blurt.log — each clip logs its length, speech level, noise floor, and dead=, the longest run of digital silence inside it. A non-zero dead= means the input dropped out mid-clip, which is what following the system default source invites; pin a device instead. Set BLURT_SAVE_CAPTURES=1 (or to a directory) to keep the raw and uploaded WAVs for every clip and hear exactly what the API got.

Checking the microphone

uv run mic_check.py [device] [seconds] records from Blurt's configured input, prints levels before and after conditioning, saves the WAV it would have uploaded, and transcribes it with the current model. Use it when transcripts come back wrong: bad audio shows up as a DC bias, clipping, or a speech level far below -30 dBFS.

Logs are written to blurt.log in your temp directory (openable from the tray menu).

Moving to another machine

uv run export.py writes a zip of just the source (no 200 MB+ .venv, no caches, no key files) to your temp directory. It's pure Python, so it runs the same on Linux and Windows — no shell needed — and it only reads, so it's safe while Blurt is recording. Unzip on the other machine and uv sync to rebuild .venv for that OS. (Or simply git clone the repo fresh.)