Home Creators Posts Import Register Favorites Logout

Content

Hello everyone,

Episode four answers the question we get more than any other about vision: how do I let my character see my screen?

- Chat Augmentations, then Voxta Utilities: Vision. It comes with Voxta and is already on; on Voxta Cloud the vision model is set up too.

- Switch on Continuous Vision and pick what she watches: webcam, screen, or both. Once a second is plenty, and the similarity filter skips frames that barely changed.

- Add Continuous Vision to your character under Options, and save. That is the whole setup.

- Then Sophie shares a shop window, asks Voxta to pick between two jackets, and switches tabs to a recipe. Voxta answers about both without anything being sent to her.

One honest note, also in the video: she reads a description of your screen, not the image itself, so small details can slip. On Voxta Cloud it uses credits for every frame it processes.

This is also the first episode with the built-in Voxta character, rewritten as a grounded companion, and the first recorded on a staged desktop, which lets the next videos show Voxta working with more than the app itself.

What would you let her look at first? Say so in the comments and the best answer goes in an episode.

Voxta is available to subscribers at: https://portal.voxta.ai/

Files

How to let your AI character see your screen (Voxta continuous vision)

The question we get most about vision: how do I let my character see my screen? This is the whole setup, start to finish, and then Voxta looks at a shop page and a recipe with Sophie and says what she thinks. Nothing is attached to the messages; she already knows what is on screen. Continuous vision takes a frame about once a second, skips frames that barely changed, and has a vision model describe the rest to your character, so small details can slip. On Voxta Cloud it uses credits for every frame it processes. Screen sharing works on a computer. 0:00 Intro 0:05 The question we get most 0:21 Chat Augmentations: Voxta Utilities: Vision 0:34 Switch on Continuous Vision, pick the screen 0:46 How often she looks, and the similarity filter 0:56 Add it to your character 1:16 Share a window and start a chat 1:39 "Green or grey? Pick one for me." 1:49 Switching tabs 2:15 One honest note Voxta is an AI platform you run and control yourself. The server is an orchestration layer over the TTS, STT, text and image generation services you choose, paid, self-hosted, or running entirely on your own hardware, with memory, scripting and scenarios built on top. It runs on Windows, Android and in the browser, and exposes an API and module system so the community can extend it rather than wait on us. Episode three, two characters in one chat: https://youtu.be/OUSMir4emRA Get Voxta (free Lite build, cloud powered): https://portal.voxta.ai/ Discord: https://discord.gg/voxta #Voxta #AIcompanion #HowTo

Comments

Trent83

Thank you for the easy explanation. I think playing The coffin of Andy and Leyey together might be interesting.