Vision BETA sample cant have CONVAI package

Unity Questions
1

Adding vision BETA to project with CONVAI Asset package = bombs
Adding CONVAI package to vision example = bombs
So, can we have them together :slight_smile:

2

Hello,

Welcome to the Convai Developer Forum!

Could you explain your question in a little more detail? What exactly are you trying to do?

3

Thank you for the quick reply. I am trying to combine Beta Vision setup and standard CONVAI sdk, they conflict and cause compiation errors, while separately each works fine

So if I want Vision beta - and I want to add Convai demos from the SDK - I cant.

And there is no Character importer in the Beta

Screenshot 2026-02-03 at 23.51.40
Screenshot 2026-02-03 at 23.51.40828Ă—271 47.5 KB

4

This is not the Vision Beta plugin. It is the new Convai SDK and is in Beta. Therefore, you cannot use it with the UAS Convai version.

5

Well, you call it BETA and it has VISION feature :slight_smile:

So, vision is standalone? It cannot be used with avatar navigation?

Screenshot 2026-02-03 at 23.54.13
Screenshot 2026-02-03 at 23.54.131485Ă—1094 179 KB

6

This is not a plugin specific to Vision. It’s simply a feature that exists only in the Beta plugin. This is highlighted in the video.

7

Got it, so vision is in beta only. Is there a tutorial on how to add the CONVAI avatar and navigation to the beta plugin?

8

The beta plugin does not support LipSync. Therefore, it currently works better in MR. So this package was created for MR.

9

I want it for MR, i want the avatar to walk in my room and tell me what’s there - case for visually impared users

10

Do you need LipSync?

11

we can add our own lipsync, for now just the avatar with nav and AI

12

You can use the beta package. This package already contains a sample scene.

13

That sample scene does not have anything else but the vision example. No avatar prefab, no avatar setup. The Beta and the SDK have very different ConvaiNPC scripts.
We really hope to not spend a ton of time evaluating the product, but it is essential for us to merge the two parts - the AI vision and the avatar with AI NLP and navigation.

14

Please make sure to carefully review the documentation and earlier replies in the thread. This Beta package is experimental, so some limitations and rough edges are expected.

A few important clarifications:

  • The current Beta sample scene is designed specifically to demonstrate new features in an MR style setup.
    That’s why:

    • There is no avatar prefab or full avatar setup included.

    • LipSync is not implemented yet in this Beta, so we did not ship a full character demo.

  • You can still add your own avatar to the scene. The sample scene does not require the avatar to be “wired” to Convai in a special way just to test Vision.

  • The Beta package and the existing SDK (Asset Store / stable plugin) are separate code paths.
    As the name suggests, the Beta is a new package and is not meant to be merged one-to-one with the old plugin.

  • For production or stable work today, the Asset Store / current stable plugin remains the recommended version. The Beta will become more complete over time, and a full version (with avatar + Vision + LipSync, etc.) will be released once it’s ready.