Your rating: None Average: 4.1 (9 votes)


Saera is a voice controlled personal assistant for mobile devices. It allows you to set alarms, send texts, read emails and more, all with the power of your voice.

You must install eSpeak first before installing this.
Application versions: 
File harbour-saera-2.0-3.armv7hl.rpm3.42 MB19/07/2015 - 02:36
File harbour-saera-2.0-4.armv7hl.rpm3.42 MB19/07/2015 - 20:19
File harbour-saera-2.0-5.armv7hl.rpm3.42 MB19/07/2015 - 23:46
File harbour-saera-2.0-6.armv7hl.rpm3.42 MB01/08/2015 - 00:54
File harbour-saera-2.1-1.armv7hl.rpm3.43 MB04/08/2015 - 00:52
File harbour-saera-2.1-2.armv7hl.rpm3.2 MB04/10/2015 - 21:53
File harbour-saera-2.1-2.armv7hl.rpm3.2 MB01/09/2019 - 15:33
File harbour-saera-2.1-3.armv7hl.rpm3.2 MB04/10/2015 - 22:12
File harbour-saera-2.2-0.armv7hl.rpm3.2 MB22/12/2015 - 02:36
File harbour-saera-2.3-0.armv7hl.rpm26.77 MB31/01/2017 - 09:27
File harbour-saera-2.3.1-0.armv7hl.rpm4.61 MB01/02/2017 - 07:46
File harbour-saera-2.3.2-0.armv7hl.rpm4.61 MB01/02/2017 - 16:36
File harbour-saera-2.3.3-0.armv7hl.rpm4.61 MB07/02/2017 - 16:39



Historyscholar's picture

How about it now?

bymoe's picture

is this project still alive ? just installed on the latest version of sailfish X and voice is not recognized .

taixzo's picture

I'm working on fixing that issue. I just got a working Xperia X last week, so I haven't had enough time to find a solution yet. (The issue is that julius' built in audio library doesn't detect the microphone. I'm trying to find a workaround that will keep the functionality as close to the original as possible.)

bymoe's picture

thanks for the quick reply and for looking into this.

Tezryuga9's picture

I tried to install the latest available repository in here on my Intex Aqua Fish running sail fish os aapajoki. I receive error every time I run it. It does not even open up the installation platform. Can anyone help me with this?

taixzo's picture

What is the error you are experiencing?

emchella's picture

i have the same problem " Initializing speech recognition" .......  and nothing more

have you some tips about?

my os version: - last jolla update.

this is the shell response:

Command harbour-saera starting...
WARNING: The 'tracker-sparql' command is no longer available.
WARNING: Use 'tracker <command> [<args>]' instead...\n
Impossibile eseguire l'interrogazione, 1.7: syntax error, expected `{'
/home/nemo/.cache/saera/musictitles.grammar has 1 rules
/home/nemo/.cache/saera/musictitles.voca    has 3 categories and 3 words
Now parsing grammar file
Now modifying grammar to minimize states[-1]
Now parsing vocabulary file
Now making nondeterministic finite automaton
Now making nondeterministic finite automaton[1/1]
Error:       undefined class "TITLE"
no .dfa or .dict file generated
Error: cannot open "/home/nemo/.cache/saera/musictitles.dfa.tmp"
Try `-help' for more information.


objectifnul's picture

Same issue here with v2.3-0 (which, by the way, has become gigantic in comparison with v2.2-0)

taixzo's picture

Released a patch that should hopefully fix these issues. Also package size is back to normal (I accidentally included flite data files in last build).

p_pahare's picture

Application is working now but No Voice Response ( Neither in output nor in Input), Only Text Commands working.

Ingvix's picture

This and also settings don't work. If swiped away from settings they return to default values.

objectifnul's picture

I confirm. Also with previous release. Additionally, manual changes in /home/nemo/.config/saera/settings.json have no effect.

Ingvix's picture

I second this.

fanfu's picture

It looks like the application cannot be installed any more. I get the following error:

Status: repo-not-available
Extra details: Download error for ''
Error code: Unrecognized error
Error message: SSL: no alternative certificate subject name matches target host name ''

Am I the only one facing the issue?

tiiagoFernandez's picture

This is amazing but, i have a few issues..

- It keeps telling me that i have no Internet connection (tried home wi-fi; street wi-fi and 4g)

- It does not understand my voice, my english is crystal clear tho

- Some things i cant manage to get them working, like sending a text

- Would like to see, maybe in about page, a full voice commands/sayings, because im not sure about them all

badpixel's picture

Can it be recompiled for Harmattan?

birdzhang's picture


     There is a question ,Martin Kolman (MartinK)  committed almost 2 years ago,and it's 1.46.02 version.The lastest is espeak-1.48 .Maybe we can upgrade it.

     So,can you build it and uploaded on your repos? I will remove espeak-1.46.02 i uploaded.


erlen's picture

Did someone succeed to use it ?
I still have the same problem when I launch it through the terminal I have this error :
Connecting to pyjulius server
ERROR: Error in loading model
And the application stay blocked with initializing speech recognition.

Wedmer's picture

Last messages before it hangs:

Connecting to pyjulius server
Try '-help' for more information

erixd's picture


I had an error where saera's UI started but without doing anything or reacting to any input. After starting it from command line, I saw that an error/exception that had something to do with $HOME/.qmf/database. This folder is generated after first start by Jollas own email application (which I did not have installed). After installing and starting that (I did not add any account), saera starts up and initializes successfully. So in my opinion email support in saera should either be optional or jolla-email should be a dependency to the harbour-saera package--and if you happen to  have a similar error, just try installing jolla-email.

erlen's picture

I launched it through the terminal and it's block under initializing speech recognition
on the terminal you have this:
Connecting to pyjulius server
ERROR: Error in loading model

arthurhkt's picture

Had anyone stuck at "Initializing speech recognition..."? Cause this is happening to me and I had no idea what should I do now :(

V10lator's picture

For me it first took ages (around 10 minutes) to initialize the music library, now it shows "Initializing speech recognition..." since around 10 minutes, too. So my guess is it just need a lot of time. Did it ever finish for you?

@taixzo Any plans to improve loading times? I mean the app is pretty useless when it needs > 20 minutes before someone is able to ask a question.

arthurhkt's picture

Nope, I left it for whole night(around 6 hour) and when I woke up, it is still showing Ini "Initializing speech recognition...", seem like a boot loop to me :(

robomike's picture

Why tries saera to fetch my GPS location? The rudimentary description here is exasperating. After stopping GPS search by killing saera the app is now aborting itself during the start ... -> deleted. 

taixzo's picture

Saera fetches GPS location so that if you ask it for the weather, for directions to a location, or for local traffic, it can do that right away without having to wait 30+ seconds for a GPS lock.

V10lator's picture

You could make an option for that (like "GPS always on" with a warning that fetching GPS location might take a long time when disabled). That way users who want to save battery life as well as privacy concerned users will be happy.

Nokius's picture

@taixzo u can add epseak/portaudio to your repo here pls see IRC logs #sailfishOS I tryed to ping u there

Slehtonen85's picture

Nothing happens when i click the microphone. I have espeak and portaudio installed.

geekonerd's picture

It's the same for me.

I've installed eSpeak for SailfishOS and portaudio, but when I launch Saera and I press the microphone icon (or I write some text) nothing appens... :(