Voxta (Virt-A-Mate Plugin)

Paid Plugins Voxta (Virt-A-Mate Plugin)

Acid Bubbles

Invaluable member
Developer
Featured Contributor
Messages
1,362
Reactions
5,415
Points
173
Website
github.com
Twitter
acidbubblesanon
Patreon
acidbubbles
Acid Bubbles submitted a new resource:

Voxta (Virt-A-Mate Plugin) - Natural speech with your AI companion

View attachment 299031

What do you receive with this resource?

- Voxta Server (Available on our Discord)
- Voxta Virt-A-Mate Plugin (This resource)
- Voxta Beta Scene (Will be created soon)

What is Voxta?


The Essence

Voxta isn't just another AI backend. It's an experience, bridging the gap between you and a realm of endless possibilities. Whether you're keen on...

Read more about this resource...
 
Question - which membership tier has access to the plugin? The companion $12 tier specifically says you get access to the plugin, but the latest post says its now available for all patrons. Does that mean its now out of beta and all companions can access it, or does that mean the user $6 tier can also access it?
 
Hey @051Kurt ! Voxta Server is the app that allows you to talk with an AI, the "backend". That's the 6$ tier. The 12$ tier includes the Virt-A-Mate plugin and scene (in addition to the Voxta Server, which is unrelated to Virt-A-Mate).

Let me know if you need more info!
 
So, maybe I didn't read everything. What is the scene for virt a mate? I watch preview and there is a sit, stand etc. voice commands which is doing our person. So, this is script commands with key words or person know their skeleton and can react to the commands like an AI? I mean..for example, if I command something like "walk 6 meters to the left" they will do this? How its work complaining with AI and skeleton of unity engine
 
@sckvair the scene is here: https://hub.virtamate.com/resources/voxta-beta-scene.40040/ but you get both when you subscribe. Actually, you get the scene, the plugin, and the server. Voxta itself is a program that can power unrestricted AI companions and assistants.

We're not trying to make _everything_ with AI. since that usually results in either unnatural animations and/or limited options. What we did instead is make a library of high quality animations that the AI can choose from; the AI part is really the brain, the personality, and deciding what to do with what their "virtual body" is able to do. We're increasing those capacities all the time, and will release new and existing scenes with other scenarios you can have fun with.

I suggest you join our Discord and check out the videos in our nsfw video sharing channel. You'll see what we mean. You can also check out those videos: https://hub.virtamate.com/threads/voxta-ai.44874/#post-132323
 
Ok thanks for the responce. Discord link will be after subscribe? I dont see it anywhere around topics
 
is it realistic/possible for the voxta team to include in your VoxtaServer app an automatic installer even for a "golden-list" services selection (a choice of what you think is the best freeware "services" for using them locally)? if yes when?
why can't you create something similar to a full local SD distro (meaning that so called "local" huge great alternative to the on-line "automatic 111" Stable Diffusion release) ?

Was it difficult or impossible to realize that just because for some of those "services" repositories it is requested a user registration (seems not) or because their related download is needing a relatively long time to be done and not possible without some specific manual user actions to skip windows 11 security blocking-alerts?

I am speaking about this manual essential procedure https://doc.voxta.ai/docs/services-overview#free-host-everything-yourself
 
@keycode We built a wizard that walks you through recommended services, and we fully handle the installation of Silero and Vosk for example. We don't (and will not) manage the installation of software like Oobabooga (it's exactly like SD automatic1111 already, actually), and cannot really impose gigabytes of software on users that may never use it because they prefer KoboldAI or running their AI on a rented GPU. There is not "one good answer", it depends on whether you have a powerful graphics card, whether you want you data to run on a server or not, etc.

So, long story short: it's already doing a lot automatically, and we made a video for Oobabooga too.

Also keep in mind, Voxta is brand new, and extremely flexible. It allows connecting multiple LLMs, and even multiple UIs. That means making "everything automatic" means "writing code for all possible use cases", and I mean.... I do my best but I can't type any faster :D

Join Discord and ask for help if you have trouble, we can definitely help you!

Finally specifically about the Windows warning... we need to buy an app signing certificate to avoid the Windows warning, and that's fairly expensive. Maybe one day we'll be large enough to warrant having an office and an official app signing certificate <3
 
@keycode We built a wizard that walks you through recommended services, and we fully handle the installation of Silero and Vosk for example. We don't (and will not) manage the installation of software like Oobabooga (it's exactly like SD automatic1111 already, actually), and cannot really impose gigabytes of software on users that may never use it because they prefer KoboldAI or running their AI on a rented GPU. There is not "one good answer", it depends on whether you have a powerful graphics card, whether you want you data to run on a server or not, etc.

So, long story short: it's already doing a lot automatically, and we made a video for Oobabooga too.

Also keep in mind, Voxta is brand new, and extremely flexible. It allows connecting multiple LLMs, and even multiple UIs. That means making "everything automatic" means "writing code for all possible use cases", and I mean.... I do my best but I can't type any faster :D

Join Discord and ask for help if you have trouble, we can definitely help you!

Finally specifically about the Windows warning... we need to buy an app signing certificate to avoid the Windows warning, and that's fairly expensive. Maybe one day we'll be large enough to warrant having an office and an official app signing certificate <3
thanks for the answer. I appreciate a lot your work with Vam (and Voxta!). I feel a lot the importance of any future progress including AI interaction with Vam . That's why I already decided to support Voxta. My best wishes from Italy. 🤞🙃
 
@sckvair Voxta already supports "all languages". I'm using quotes here because it depends on the service you use. For example, Deepgram supports most languages for speech to text, and ElevenLabs supports all of them, Silero supports a bunch, etc. For the large language model (the brain), OpenAI does not allow erotic role play but has great language support, otherwise there are LLMs for pretty much all languages too. Voxta itself (the app) is only in English but you can create prompt templates in other languages, too.
 
So the unrestricted AI version may be only local-installed? This is what i hear on video. Ask becuase of I was use some local AI in the past and they work slow at 7B models with my 8GB VRAM =(
 
Hello, I joined your Patreon (the middle tier), and went to the discord, but cannot download the plugin.

It says this:

- Download Voxta Server (for speaking with the AI and the backend for Virt-A-Mate) here: no access
- Download Voxta Virt-A-Mate Plugin (for using with Virt-A-Mate) here: no access
- Download the Beta scene for Virt-A-Mate here: no access

Am I missing something? How can I download the plugin? Thanks!
 
Finally, it's not just come here and F me... you can actually have a personality, and have to talk to them, (given you don't set their mind and memory to really dirty), to get a reward. You can have multiple scenes, animations and actions, etc... give at a few more years of hardware development and improved AR, it'll be like having that live in companion, you'll see him/her when put your glasses on, and yeah... wow...
 
What type of performance draw does the ai engine need locally? Does it fully support hardware acceleration? Wondering if its fine on my 3070 on top of vam or should I run the server on another amd computer I have that has more limited ai acceleration.
 
I'm still imagining that if one day, AI can train mocap data very well and then apply it to VAM, whether there will be a completely random, coherent, natural, real and unpredictable character animation. She can pose continuously, sit down and stand up, and also can twist and dance simply. I don't understand why AI has not trained mocap data and combined it with VAM. In theory, mocap data is controllable variables, and its regularity and cleanliness are very suitable for AI training requirements. I don't know why no one has made it applied to the VAM characters.😭
 
@sckvair Many services are unrestricted, such as NovelAI (though it has trouble with instructions and logic, but the voice and writing is nice) and OpenRouter, and soon Voxta Cloud but the best options are always self hosted. With 8GB you can't run everything locally though, at least not "good" options like xtts, 7B GPTQ models etc. Requirements keep getting smaller though :)
 
@huiyu914 Mostly because building AI is hard, building applications with AI is also hard (at least to do it well) and everything's being developed at a dizzying speed! Everything will come, as people build them :)
i have used voxta to talk with AI, it is amzing! (y)
but when i only want to sex, she's talk is not important, the mocap、expression is the key point.
look forward the AI+mocap,AI+expression ,
wish voxta can support it someday haha ;)
 
Back
Top Bottom