# AI Models ## Language Models The AI model that you chat with inside a chat are Large Language Models. There are currently three different LLMs available. ### **Bacchus** A lightweight model which is fast, expressive, but may not be very intelligent. Bacchus is great for casual chats, punchy replies, and expressive behavior. It may not perform as well for complex stories, but Bacchus can write with a rawness that the other models may sometimes lack. {% hint style="success" %} Hint: this model does especially well with violence and NSFW. {% endhint %} ### **Jupiter** A solid, flexible model. Jupiter is a jack-of-all-trades, capable of writing everything from simple chats, complex stories, and everything in between. Jupiter portrays characters well, is capable of moving the story forward, and can handle most character types with a good degree of nuance. ### **Infinity** A large, intelligent model that can handle complex stories and detailed plots. Nuanced, good at handling complex themes, supplementary characters, following prompts and everything else. Infinity generates replies slower than the lighter weight models but what it lacks in speed it makes up for in capabilities. {% hint style="info" %} All three models have the same context limits based on subscription level. Bacchus is an 8b model, Jupiter is a 70b model and Infinity is a \~300b model. {% endhint %} ## Image & Video Models Xoul.AI uses various image and video models which you can find more information on in [AI Images, Video & More](/xoul.ai-official-guide/navigation-and-information/navigation-and-interfaces/content-creation/ai-images-video-and-more.md). **To summarize:** * **Image Models:** Flux, GLITCH, Float, Nano Banana, Ideogram, Seedream, GPT Image and Flux Kontext. * Each of these models have a variable cost, determined by the exact settings being used, but most Cell prices range from 1-9 Cells per generation. * **Video Models:** VEO and Seedance. * Models have a variable cost, with Seedance being 10-150 Cells per video depending on settings and VEO being the more expensive model at 320-600 Cells per video. ## Voices Xoul.AI uses ElevenLabs to power voice features across the platform. Users can generate or clone images for 30 Cells, use voice narration in chat, create talking avatars by combining voices with static images (price varies by length), and use the Call feature to chat with a character. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://xoul-ai-official-documentation.gitbook.io/xoul.ai-official-guide/navigation-and-information/ai-models.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.