I'm working on a robotics project that requires speech recognition. I want the finished product to be portable, so it will almost certainly use a Raspberry Pi. Is there any speech recognition software (or a speech recognition library) that runs efficiently on a Pi?


  • Is for the English language.
  • Runs efficiently on a Pi.
  • Is...
    • ...able to be modified at the source code level (to run the robot), or
    • ...a library, so I can make my own program.
  • Is free or inexpensive.
  • Does not require internet access.

1 Answer 1


I am a developer of CMUSphinx, an open source speech recognition toolkit. The main features of CMUSphinx for robot control and Raspberry Pi are:

  • Accurate models are available for US English
  • Good models are available for French, German and Spanish
  • Any language can be supported
  • Efficient decoding on low-resource hardware
  • Noise-robust speech recognition
  • Continuous listening mode with keyword activation
  • Support for various languages like Python or Ruby

You can learn more about CMUSphinx from the wiki. If you have questions about CMUSphinx, feel free to ask on forums.

  • Do you have any feedback to provide about your experience of using CMUSphinx? Mar 5, 2014 at 9:48
  • 1
    @Fractaliste, As the developer I can guarantee you that CMUSphinx can exceed your needs on speech recognition. If you have some specific question, feel free to ask. Mar 5, 2014 at 14:52
  • Hi Nikolay, could you contribute an answer to this question? softwarerecs.stackexchange.com/questions/17124/… Mar 15, 2015 at 23:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.