AI Chatbot Helper [FEEDBACK NEEDED]

We’re excited to introduce the Thunkable Chatbot, your always-on, AI-powered app-building helper, now available to all Thunkers!
Whether you’re just getting started or you’re a seasoned creator, the chatbot is here to support you 24/7. From answering your development questions to guiding you through tricky steps, it’s like having an expert by your side, no Googling or guessing required.

Why you’ll love it:

  • Instant answers to your app development questions
  • Step-by-step explanations and tips
  • A faster, smoother way to bring your ideas to life

How to use it:

  1. Open your project

  2. Go to the Block tab or the Design tab

  3. Click the AI icon


  4. Write your question

We’re actively collecting feedback to improve your experience.
Try it out and let us know what you think!

Learn more about:

AI Chatbot Helper

How do tokens work?

How to get tokens?

2 Likes

@tatiang @krish05 @sketch @martin.carl.zimmerl8 @annaradnor70pzj4mu @grant.mccallum @22kifayatun8ne @anandavardhana57zwwl @iffsydl78qiz @maurizio.polverini89

Looking forward to hearing your thoughts!

  1. The ? box appears only when we open an project. That is not right.
  2. It should appear when we just open My Projects too.
  3. As a mentor I have 4 screens open and assume I have 1 project open on each. When I click on ? on one of them all the 4 come up with the Chatbot screen. That is very bad and irritating. So please restrict the Chatbot screen only for the project it was opened and please ensure that all future conversation appear only when that project is opened,
  4. When the Chatbot is opened it covers the Share menu. Assume I have typed an issue and want to share a link to the program then I have to close the Chatbot to access that. That is also very unfriendly. Best thing is to make it movable within the screens and it is not stuck. Right now it is stuck at the right top corner of the project. Ideally I should be able to open it and drag it to a different monitor(Remember I have 4 monitors) then life becomes easy for us to look the program and explain our issues
  5. Earlier we made the request in the Chatbot and later all conversations was via email. Nowadays most of the time the Email is empty other than saying the bug has been filed etc … So please make sure you revert back to the old format so that the very first time we file the bug report via the Chatbot and all subsequent conversation is with email. The reason is I dont want to open Thunkable/The project to converse. Assume I am outside and would like to converse then using a Mobile becomes just impossible. On the other hand using the mobile to converse using email becomes much simpler (Senior Citizen friendly). So only the initial conversation when we file let it be on Chatbot later on all in email.
  6. I just got a ticket closure and with that I got this request

    Bug1: I got two of them
    Bug 2: Both had been greyed/disabled so I could not click them
    So closed it without rating
  7. Then when I relogged in and clicked the help button I got this:

    I guess it is remembering that I have not rated the earlier one :slight_smile:
    Sorry for this series of bug report but you asked for it and like a deligent user shared it!!
    Bottom line, these are all small issues I love working with Thunkable and my 8 mentees are having lot of fun and learning a lot. A BIG thanks for all you folks. Keep up the good work!!
1 Like

Hi @anandavardhana57zwwl
Thanks for your feedback. It helps a lot.
Just to clarify that the chat in your last screenshots is from Intercom chat and not AI Chatbot Helper.

1 Like

Hi @ioannis

Oops sorry missed it. BTW one more bug :slight_smile: on my previous comment is, nowhere it says Intercom Chat. In the email it says Thunkbot. So please fix that issue also. :slight_smile:
Now coming to AI Chatbot.

  1. AI Chatbot does not show up in Organizational projects. That is all of our work happens just to chat we have to switch to My Projects. Please fix it
  2. In My Projects it does show up as you say. However if I may suggest when we hover over it, make it to display its name then users will know that exists and will click on it. When we are talking of Hovering please do that for the “xxxx tokens left” thing which started coming recently. I thought it was bug :slight_smile: initially.
  3. I just clicked it, it does not say AI Chatbot helper but says StyleAI. That is a cute name, I like it but is not what you said. Am I missing anything pardon my ignorance
  4. This is the question I asked it:

How can I convert a Image to a base64 image so that I can save it to a Google Drive?

To convert an image to a Base64 string in Thunkable, you can use the following approach:

  1. Use the Camera or Media DB Component:

    • If you are capturing an image using the Camera or fetching from Media DB, these components can provide the image data.
  2. Use the “Base64 encode” block:

    • In Thunkable, use the “Base64 encode” block to convert the image to a Base64 string. This block takes the image data as input and outputs the Base64 encoded string.
  3. Save to Google Drive:

    • Once you have the Base64 string, you can use an API call or Google Drive’s API to upload this string as a file to Google Drive.

Here’s a basic example of how you might set this up with blocks:

  • Capture or select an image.
  • Use a block to convert that image to a Base64 string.
  • Send the string to Google Drive using an API integration component.

You will need to ensure you have the right permissions and set up to interact with Google Drive’s API to upload the file.


My Comment: If I ask ChatGPT that also says the same thing. What both are missing is pinpointing as to how to get access to “Base64 encode” block. ChatGPT might not have access to the internals’ of Thunkable whereas StyleAI has and it should give more precise answer.


Then I asked this question:
Can you show me how to access the Base64 encode block? Under what category will it come? Do we have to install something else for it show up?
This is what it answered:


In Thunkable, the “Base64 encode” block is typically part of the standard functions available to you without needing additional installations. However, the exact location or availability might depend on updates or the specific components you have added to your project.

Here’s a general guide on how you might find and use the Base64 encode block:

  1. Open Blocks Editor:

    • Go to the Blocks section of your Thunkable project.
  2. Look for Text or Utilities Section:

    • The Base64 encode functionality might be found under text manipulation or utilities functions. Check under the “Text” category or any category that involves data manipulation.
      My Comment: StyleAI is Thunkable product which should have access to the latest guts of Thunkable. It can see what version of Thunkable the user is running and tell them in what version and exactly where to find it. Instead of giving soem vague answers.
  3. Check for Component-Specific Blocks:

    • If you are working with a component like the Camera, Media DB, or any image-related component, examine if there are blocks related to data encoding or handling that component’s data.

If you do not see a Base64 encode block directly available, you may need to create a function or use a workaround by integrating with a web API or using JavaScript through a Web Viewer to handle Base64 encoding.

Remember, Thunkable frequently updates their platform, so the exact location or availability of specific blocks might change. Always refer to the latest documentation or community forums if you encounter any issues. If you do not find it, you might want to check the Thunkable Community for any recent changes or solutions provided by other users.


My Last Comment on this: I had asked one of you and you had told me Thunkable does not support Base64 convertor module yet. ,StyleAI also should be aware of it and give a brief to the point answer not same vague suggestions. As applicable to Thunkable StyleAI should do a better job than ChatGPT

I’d love to have a play with this but the icon isn’t appearing for me. In any case, I’d move the icon to the help section in the top right with your other support options as this feels a lot more intuitive.

I am having Advanced Version of Thunkable, just FYI. Please let us know what version of Thunkable you have? As I said earlier the StyleAI does not show up in Organizational version of Thunkable too…

Thanks for your feedback.
@anandavardhana57zwwl I am sorry that it was frustrating. I will flag these to the team.
@grant.mccallum This is happening because the AI features are not available to creators with the old plans.

This is my question: How do I convert a photo image to a base64 encoded image?

ChatGPT and your AI Chatbot Helper repeated the same below message:

My point is: Thunkable does not support the blocks they mention. ChatGPT is an external tool does not have access to the internal guts of Thunkable and it is telling us the info it garnished from internet. Some older version of Thunkable did support it hence it spits out that info. That is excusable.

My Bug report: AI ChatBot is a Thunkable tool and should be aware of obsoleted blocks, it should have access to the internal guts of Thunkable, It should know better than repeat what ChatGPT said. As applicable to Thunkable it should have the latest info, else what is the point, if it repeats ChatGPT? Right?

My Requests are

  1. Fix AI Chatbot helper

  2. Provide the old blocks which aided in base64 conversion.
    ********************* Message from ChatGPT and AI Chatbot helper*******************
    To convert a photo image to a Base64 encoded string using Thunkable, you can follow these steps with the specific blocks:

  3. Capture or Select the Image:

  • Use the Camera or Photo Library component to capture or select an image.
  1. Set Image on Canvas:
  • Add a Canvas component to your screen.
  • Use the block to set the background image of the Canvas to the captured or selected image. This can typically be done using a block like canvas.setBackgroundImage.
  1. Get Base64 String from Canvas:
  • Use the block canvas.getBase64Image to retrieve the Base64 encoded string of the image drawn onto the Canvas.
  1. Use the Base64 String:
  • Once you have the Base64 string, you can use it to send to your OpenAI API.

One more bug on AI Chatbot helper. If you scroll up, earlier I had filed a bug report saying when we click the ? and request tech help from live people the help box covers the “share link” generating menu item. Someone obviously fixed it and now the ? help screen gets displayed at the bottom, so that the Share menu is visible. But the new issue is it covers the AI Chatbot helper :slight_smile: :slight_smile: The easiest fix for both the AI and Live help screen is to make it movable.