Google Conversational Actions Sunset



Hey everyone,

Google decided that they are sunsetting Google Conversational Actions (which means all custom conversational apps built on top of the Google Assistant platform). You can learn more here:

What this means:

  • After June 13, 2023 (in about a year), Google will remove all Google Conversational Actions from the Google Assistant “marketplace” -> existing Actions won’t work anymore and it won’t be possible to publish new ones
  • This includes all custom Actions, there is no way to circumvent this by using the old Dialogflow Assistant integration
  • The only types of Actions that are going to work are so called “App Actions” that work in a different way and are tied to an Android app

Unfortunately, this means we won’t be able to build Google Actions with Jovo anymore. This is a bummer because a lot of people (including us) have invested time and money into the Google Assistant ecosystem. However, we learned a lot while integrating with this platform and took a lot of ideas from it into the development of Jovo v4.

Maybe it’s time for you to finally move towards building a custom solution? Or looking for other interesting platforms to deploy your Jovo apps to? You can find everything in the Jovo Marketplace:

Let us know if you have any questions!


As a voice, or STT interface, I mainly used and preferred Google Actions because it provided the full user utterance, something that Alexa does not do. Which other systems are available that would provide this same, or similar functionality?


Our Web Platform together with an ASR integration like Amazon Lex could work for you. Or are you looking for a smart speaker platform?


I am looking for a mobile enabled thing that I can talk to while wearing earbuds, as was building around Google Assistant. I find it hard to believe that in these times of rising conversational agents, Google would be dropping out and removing its product, unless it is trying to force users down a different path or to use another one of their products, perhaps even something new. None of their posts offer any reasons why or any alternatives, I find the whole thing a bit odd.


Google is trying to push Google Home users toward an Android-code based system, as I understand it based on what I’ve seen on the web.

“Odd”, unfortunately, usually means “there’s a marketing reason we don’t have insight into”, and there isn’t a lot one can do about that except complain directly to the company and hope they will eventually explain.

My unsupported guess is that they’ve decided stand-alone conversational devices aren’t a market they can compete effectively in, given Alexa’s market share. If so I’d expect to see them focus on things based on phones, or integrated into consumer products like TV sets… and to put their emphasis on doing a much smaller set of things well rather than trying to be a general conversation environment.

If/when they say something coherent, I’ll consider looking at them again; for now it’s simply not worth my effort to chase them.


Thanks Joe. So Google Assistant will only talk to bots built in Android, that’s Java isn’t it? Like you say, no point in wasting energy chasing reasons, better to put that energy into learning something new that will stick around. Jan suggested Lex, but are there any other options to consider, because now would be the time to consider them right? I really only need the thinnest client running on a mobile phone, and passing all the text into Jovo where nlp.js is doing the NLU. Can’t use a browser because most don’t enable STT, or TTS or both. Desktop Chrome is the only one that does Webspeech. Do I need to build an Android App?


Haven’t investigated. Not currently planning to. All I can do is wish you luck and remind you that Google theoretically has support resources.