A few questions

While searching for an open source equivalent of Alexa or Google, I found Mycroft and it looks very interesting!

1- I am wondering what hardware I could use for the following applications:
1a) Standalone MP3 player from a sd card (or usb stick) with decent speakers quality.
1b) Robot (walking or rolling) with camera (for object/face/obstacle recognition), video for displaying robot emotions and some sort of hand or grip to grab or move stuff around.
1c) Controlling house automation using zigbee, zwave or some other low power transmission. Probably not WIFI since from what I read, it takes a lot of power.

2- Is there a way that Mycroft can run standalone? Speech to text already exists on standalone PC for many years. Or if it’s a question of too much resources, perhaps connect to my own local server (or app?), without internet connection. I prefer not using cloud or internet servers, because of privacy, dependability and control over what is going on.

3- For programming some skills or changing the AI behavior, what language(s) can be used?

4- Any good tutorials recommended for all the above?


  1. Depends on the hardware you run on, a pi3 wouldn’t be sufficient for b, possibly c (someone else may have a better answer on that), but fine for a.

  2. Yes, the selene backend or the personal backend can do this.

  3. Python is primarily used with mycroft, but you could theoretically use anything if so motivated. R also gets some AI usage.

  4. There’s a magic mirror and talking teddy bear thread here on the forum I don’t have links handy for.

Thanks for taking the time to answer!

I am still wondering which hardware I could use though.

What do you have, and what are you hoping to run on it besides mycroft?