Nerd-dictation, hackable speech to text on Linux

Nerd-dictation, hackable speech to text on Linux

Offline Speech to Text for Desktop Linux. – See demo video.

This is a utility that provides simple access speech to text for using in Linux
without being tied to a desktop environment.

This is a single file Python script with minimal dependencies.
User configuration lets you manipulate text using Python string operations.
Zero Overhead
As this relies on manual activation there are no background processes.

Dictation is accessed manually with begin/end commands.

This uses the excellent vosk-api.


It is suggested to bind begin/end/cancel to shortcut keys.

For details on how this can be used, see:
nerd-dictation --help and nerd-dictation begin --help.


Specific features include:

Numbers as Digits

Optional conversion from numbers to digits.

So Three million five hundred and sixty second becomes 3,000,562nd.

A series of numbers (such as reciting a phone number) is also supported.

So Two four six eight becomes 2,468.

Time Out
Optionally end speech to text early when no speech is detected for a given number of seconds.
(without an explicit call to end which is otherwise required).
Output Type
Output can simulate keystroke events (default) or simply print to the standard output.
User Configuration Script
User configuration is just a Python script which can be used to manipulate text using Python’s full feature set.

See nerd-dictation begin --help for details on how to access these options.


  • Python 3.
  • The VOSK-API.
  • parec command (for recording from pulse-audio).
  • xdotool command to simulate keyboard input.


pip3 install vosk
git clone
cd nerd-dictation
mv vosk-model-small-en-us-0.15 model

To test dictation:

./nerd-dictation begin --vosk-model-dir=./model &
# Start speaking.
./nerd-dictation end
  • Reminder that it’s up to you to bind begin/end/cancel to actions you can easily access (typically key shortcuts).

  • To avoid having to pass the --vosk-model-dir argument, copy the model to the default path:

    mkdir -p ~/.config/nerd-dictation
    mv ./model ~/.config/nerd-dictation


Once this is working properly you may wish to download one of the larger language models for more accurate dictation.
They are available here.


This is an example of a trivial configuration file which simply makes the input text uppercase.

# ~/.config/nerd-dictation/
def nerd_dictation_process(text):
    return text.upper()

A more comprehensive configuration is included in the examples/ directory.


  • The processing function can be used to implement your own actions using keywords of your choice.
    Simply return a blank string if you have implemented your own text handling.
  • Context sensitive actions can be implemented using command line utilities to access the active window.


Local Configuration
Language Model


Note that --vosk-model-dir=PATH can be used to override the default.

Command Line Arguments

Output of nerd-dictation --help


nerd-dictation [-h]  ...

This is a utility that activates text to speech in Linux.
While it could use any system currently it uses the VOSK-API.

positional arguments:

begin: Begin dictation.
end: End dictation.
cancel: Cancel dictation.
-h, --help show this help message and exit

Subcommand: begin


nerd-dictation begin [-h] [--cookie FILE_PATH] [--vosk-model-dir DIR]
                     [--pulse-device-name IDENTIFIER]
                     [--sample-rate HZ] [--defer-output] [--continuous]
                     [--timeout SECONDS] [--idle-time SECONDS]
                     [--delay-exit SECONDS]
                     [--punctuate-from-previous-timeout SECONDS]
                     [--full-sentence] [--numbers-as-digits]
                     [--numbers-use-separator] [--output OUTPUT_METHOD]
                     [- ...]

This creates the directory used to store internal data, so other commands such as sync can be performed.

-h, --help show this help message and exit
--cookie FILE_PATH
  Location for writing a temporary cookie (this file is monitored to begin/end dictation).
--vosk-model-dir DIR
  Path to the VOSK model, see:
--pulse-device-name IDENTIFIER
  The name of the pulse-audio device to use for recording.
See the output of “pactl list sources” to find device names (using the identifier following “Name:”).
--sample-rate HZ
  The sample rate to use for recording (in Hz).
Defaults to 44100.

When enabled, output is deferred until exiting.

This prevents text being typed during speech (implied with --output=STDOUT)

--continuous Enable this option, when you intend to keep the dictation process enabled for extended periods of time.
without this enabled, the entirety of this dictation session will be processed on every update.
Only used when --defer-output is disabled.
--timeout SECONDS
  Time out recording when no speech is processed for the time in seconds.
This can be used to avoid having to explicitly exit (zero disables).
--idle-time SECONDS
  Time to idle between processing audio from the recording.
Setting to zero is the most responsive at the cost of high CPU usage.
The default value is 0.1 (processing 10 times a second), which is quite responsive in practice
(the maximum value is clamped to 0.5)
--delay-exit SECONDS
  The time to continue running after an exit request.
this can be useful so “push to talk” setups can be released while you finish speaking
(zero disables).
--punctuate-from-previous-timeout SECONDS
  The time-out in seconds for detecting the state of dictation from the previous recording, this can be useful so punctuation it is added before entering the dictation(zero disables).
  Capitalize the first character.
This is also used to add either a comma or a full stop when dictation is performed under the
--punctuate-from-previous-timeout value.
  Convert numbers into digits instead of using whole words.
  Use a comma separators for numbers.

Method used to at put the result of speech to text.

  • SIMULATE_INPUT simulate keystrokes (default).
  • STDOUT print the result to the standard output.
    Be sure only to handle text from the standard output
    as the standard error may be used for reporting any problems that occur.
- … End argument parsing.
This can be used for user defined arguments which configuration scripts may read from the sys.argv.

Subcommand: end


nerd-dictation end [-h] [--cookie FILE_PATH]

This ends dictation, causing the text to be typed in.

-h, --help show this help message and exit
--cookie FILE_PATH
  Location for writing a temporary cookie (this file is monitored to begin/end dictation).

Subcommand: cancel


nerd-dictation cancel [-h] [--cookie FILE_PATH]

This cancels dictation.

-h, --help show this help message and exit
--cookie FILE_PATH
  Location for writing a temporary cookie (this file is monitored to begin/end dictation).


  • Typing in results will never press enter/return.
  • Pulse audio is used for recording.
  • Recording and speech to text a performed in parallel.


Store the result of speech to text as a variable in the shell:

SPEECH="$(nerd-dictation begin --timeout=1.0 --output=STDOUT)"

Example Configurations

These are example configurations you may use as a reference.

Other Software

  • Elograf – nerd-dictation GUI front-end that runs as a tray icon.


  • Text from VOSK is all lower-case,
    while the user configuration can be used to set the case of common words like I this isn’t very convenient
    (see the example configuration for details).

  • For some users the delay in start up may be noticeable on systems with slower hard disks
    especially when running for the 1st time (a cold start).

    This is a limitation with the choice not to use a service that runs in the background.
    Recording begins before any the speech-to-text components are loaded to mitigate this problem.

Further Work

  • And a general solution to capitalize words (proper nouns for example).
  • Wayland support (this should be quite simple to support and mainly relies on a replacement for xdotool).
  • Add a for easy installation on uses systems.
  • Possibly other speech to text engines (only if they provide some significant benefits).
  • Possibly support Windows & macOS.

NOW WITH OVER +8500 USERS. people can Join Knowasiak for free. Sign up on
Read More



“Simplicity, patience, compassion.
These three are your greatest treasures.
Simple in actions and thoughts, you return to the source of being.
Patient with both friends and enemies,
you accord with the way things are.
Compassionate toward yourself,
you reconcile all beings in the world.”
― Lao Tzu, Tao Te Ching