Vanessa LLM

Personal, Portable, Offline; Vanessa is here to allow you to run Large Language Models straight from your device, No internet required!
Privacy is everything! Vanessa will not save your conversations anywhere, not even on your device to ensure full privacy and security. The LLM will know the context of the chat until you restart the app, which will then load up a new conversation with no memory of the previous one.
It is recommended to close all background apps due to the apps hefty RAM requirement otherwise, it will crash the app/system
Vanessa is offered in three models:
--Mistral--
This 7B parameter model will only be available to some device models due to the hardware requirements. This model is slower than the other model but is way less prone to mistakes than TinyLlama. If you have a compatible device with 8GB ram or above, you will be prompted to optionally download the model.
--Rocket--
This 3B parameter model is the perfect middle ground for speed and quality providing decent quality generation without sacrificing to much speed. This model will only be available on devices with 6GB ram and greater.
--TinyLlama--
This is a 1.1B parameter model and was chosen due to the less strict hardware requirements and will be available on all phone models this app supports. This model is noticeably faster at generation than Mistral and Rocket but is more prone to errors and mistake. Only available on devices with 4GB ram.
Restart the app if anything goes wrong.
TinyLlama, (the default model on this app) is prone to errors and hallucinations.
It is important to check important information.
This app is NOT supported on:
iOS 16 or older operating system version is not supported.
--iPhone--
- iPhone 4
- iPhone 5
- iPhone 6
- iPhone 7
- iPhone 8
- iPhone X
- iPhone SE
- iPhone SE2
--iPad--
- iPad Gen 1, 2, 3, 4, 5, 6, 7, 8, 9
- iPad Mini Gen 1, 2, 3, 4, 5
- iPad Air Gen 1, 2, 3
- iPad Pro 9Inch
I, Noah Kim, will not be held responsible or responsible for anything that the LLM's produce.