@tishko_s You need to drag the plugin element on the page for the actions to become available 
1 Like
Thank you. I made it work.
How can I make it work if I want to create 2 buttons for “Start” and “Stop” and without getting error when clicking “Start” twice by mistake?
On your Start button workflow put the condition Real-TimeSpeechToText A's Is Listening is no
Then putReal-TimeSpeechToText A's Is Listening is yes on the Stop button workflow
1 Like
Amazing, works great now. I am only using 1 button for Start/Stop! Thank alot for the support!
2 Likes
It would be awesome if this can stop or pause on voice silence and if we can control that time. Right now, I think it stops listening if there is a gap of 3 seconds, we need to be able to control that
This is entirely controlled by the browser/OS, unfortunately there isn’t much we can do for now 
Hey i am really interested in buying that plugin but i guess speech to text is not that accurate
This plugin leverages onboard speech-to-text of your browser. Therefore, accuracy is directly dependent on this implementation.
Some browser may also perform remote recognition while connected to internet, which might lead to accuracy variation. Some others are leveraging onboard devices when available.
Moreover, accuracy greatly varies depending on languages.
Long story short, accuracy isn’t something I have control on, besides making sure the correct language is selected of course.
If you need a more robust solution, you might want to try Azure AI - Endless Speech-to-Text Plugin | Bubble or Azure AI - Transcribe Conversation Plugin | Bubble
Tried using this and while it does translate speech to text, the issue is how the app stores the recorded values. Basically, it just keeps logging one large text string. This means that to get the last sentence, you need to keep regex-ing the string waiting for the tool to put in a “.”
This is unusable for anything beyond simple command detection. We are trying to use it to facilitate a conversation between the user and AI so as you can imagine this global “Value” just keeps getting bigger and bigger so the regex keeps taking longer and longer to peruse the string and find the last sentence.
Can this (as an option) return sentences (which I know it does detect because it places the period at the end of the sentence) so that the event Value Changed is triggered when the sentence is completed rather than on every word which is currently happening?
I understand the difficulty in what’s being done here, but if you are going to charge for a plugin like this, it needs to be able to do the basics.
Dear @sanastasi ,
Thanks for your message.
To clarify, this plugin does not include its own speech recognition engine.
Instead, it relies on the Speech Recognition APIs that are natively provided by browser vendors. As a result, the quality and range of features can vary significantly depending on the browser.
For example:
- Feature coverage: Some browsers support advanced options such as interim results, continuous listening, or multiple languages more fully than others.
- Platform differences: The same browser may behave differently depending on whether it’s running on desktop, mobile, or a device with on-chip speech recognition (e.g., certain Android phones).
- Vendor implementations: Since the API specification leaves room for interpretation, Chrome, Edge, and Safari each implement it slightly differently, leading to inconsistencies in performance and reliability.
- Sentence detection & punctuation: The APIs generally do not handle sentence boundary detection, and in some cases they do not provide automatic punctuation at all. This means developers cannot rely on consistent formatting of recognized speech.
In short, the plugin’s behavior ultimately depends on the capabilities and limitations of the user’s browser rather than on the plugin itself.
I would also mention that this plugin is used to number of apps and fits perfectly their use-cases, which may certainly be less demanding that yours.
To better suit your need, I would recommend Azure AI - Endless Speech-to-Text Plugin | Bubble, or moving directly to a native audio AI such as OpenAI - GPT Realtime Conversation Plugin | Bubble