​
Model lineup

ModelDescription
Vivian (vivian-llama3.1-70b-1.0-fp8)A 70B parameter model based on Llama 3.1. A general purpose conversational model built specifically for relationship building and chatting. It has a context window of 8k tokens.
Xavier (xavier-r1)Our reasoning model, also a 70B parameter model built on top of Llama 3.1 that excels at emotional reasoning. This model outputs thinking or reasoning tokens so it’s generally slower, supports 8k tokens and is currently under limited release.

​
Model ID aliases and snapshots

In our API, you’ll use model IDs to specify which model you want to use. The model IDs correspond to specific versions of our models. 
import OpenAI from "openai";

const openai = new OpenAI({
  baseURL: "https://api.animusai.co/v2",
  apiKey: process.env.ANIMUS_API_KEY,
});

const completion = await openai.chat.completions.create({
  model: "vivian-llama3.1-70b-1.0-fp8", // Using the Vivian model
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    {
      role: "user",
      content: "What have you been up to lately?",
    },
  ],
});

console.log(completion.choices[0].message);

​
Choosing the right model

When deciding which model to use for your application, consider the following factors:
  • Task type: Choose Vivian for general conversation and Xavier for complex reasoning tasks.
  • Performance needs: Consider response time requirements for your application.
  • Token usage: Be aware of token consumption differences between models.
  • Cost considerations: Different models have different pricing structures.

Vivian

Learn more about our general purpose conversational model

Xavier

Explore our emotional reasoning model with thinking capabilities