Character is not able to perform more than one action, when using the Quest 3
What happened?
I created a new scene with number of actions and known obkects. when I test in the unity editor my character is able to perform a number of actions (one by one) when I ask her to, like walk to window, or point to painting, and walk to door.
HJowever when I try to perform the same actions when compiled to Met Quest 3 and running on the quest, then all seem to work fine, until the first action is done, like if I ask her to walk to de door, she wil ldo and then can’t walk anymore, If I ask here why she tells me that this is currently not possible or she is not able to do that right now
I checked the relevant documentation, resources, and similar forum posts before posting.
on
I confirm that I did not include API keys, tokens, passwords, secrets, private character data, private conversation logs, or customer data.
You can find guides online for collecting logs from Meta Quest devices, or you can ask an AI assistant for step by step instructions based on your setup.
Please collect the relevant Quest/Android logs and share them with us so we can review the issue further.
I was able to do some more testing.
I noticed that if I use the DEMO Character ID: 51a5cb20-01e7-11f1-a307-42010a7be027 Sofia,
Than the system is much more responsive and executes al my actions, it does have some delays every now and then, however it does execute most of my commands, only stopped once telling me it could not perform an action. If I use my own Character ID: dd6e9ea8-23fc-11f0-9ce6-42010a7be01a Sandra, then I can only execute one action (only once it performed two actions) and then stops telling me it can not perform that action (Log attached)
What are the differences between the two character ID’s that can cause this behavior?
Hi Kaan,
There are so many to choose from, how do I know what is a good one?
Could you tell me what LLM model is used for DEMO Character ID: 51a5cb20-01e7-11f1-a307-42010a7be027 Sofia.
That could be a starter LLM for me.
Is there a way to understand the differences between the models?
How do I know what is a good one, for my purpose?
You can choose a model based on the Intelligence and Latency values shown in the LLM selection screen.
For the demo character Sofia with Character ID 51a5cb20-01e7-11f1-a307-42010a7be027, the model being used is Gemini 3.1 Flash Lite.
In general, if you need better reasoning and more reliable behavior, choose a model with higher intelligence. If faster response time is more important, choose a model with lower latency.
You can also improve Action triggering by adding clear instructions to the character description so the character understands when and how to use the configured Actions.