Create: Selecting right Models

Moemate offers a wide variety of Large Language Models (LLMs). How do you choose the perfect model for your character?

It all depends on your use-case. Is it your assistant? Is it for roleplay? Maybe you’re learning a language? Whatever the case may be, we’ll help you choose the right LLM.

This guide will give a brief breakdown of the various LLMs available on Moemate and what unique qualities they each have to offer.

Choose from the following LLMs:

  1. GPT 4

  2. GPT 4 Turbo

  3. GPT 3.5 Turbo

  4. GPT4o

  5. GPT4o mini

  6. Claude Instant V1.2

  7. Claude V2

  8. Claude V2.1

  9. Claude V3 Sonnet

  10. Claude V3.5 Sonnet

  11. Claude V3 Haiku

  12. Moemate LLM

GPT 4

GPT4 is the best and smartest model available. It’s a proprietary model, available through the OpenAI API. It offers a 6k context length (compared to the 4k length available on other similar models). It’s an extremely intelligent model capable of in-depth discussions on a wide range of topics, even in multiple languages. However, since GPT4 is not an open source model, it does censor itself on particular NSFW topics.

Overall, GPT4 is the perfect all-rounder, censoring notwithstanding.

GPT 4 Turbo

The latest GPT-4 model with improved instruction following, JSON mode, reproducible outputs. Model may exhibit some instability. A stable model is being released in the coming weeks.

GPT3.5 Turbo

Similar on most fronts to GPT4, GPT3.5 isn’t quite as smart, but is much faster.

GPT 4o

GPT 4o is the most advanced model that's faster and cheaper than GPT-4 Turbo.

GPT 4o mini

GPT-4o mini is the most recent OpenAI model, a small model with superior textual intelligence and multimodal reasoning that enables a broad range of tasks with its low cost and latency.

Claude Instant V1.2

Faster than its big brother Claude v2, however poor at using skills or modules.

Claude V2

Claude v2 is another very popular model known for its creativity during roleplay conversations. Claude LLMs have context lengths of 6k-7k, which enables their conversations to stay consistent for much longer. We’d suggest you use Claude v2 for roleplay characters as it tends to be the most creative. Although keep in mind that Claude - like GPT - censors, so no NSFW.

Claude V2.1

Claude 2.1 has made significant gains in honesty, with a 2x decrease in false statements compared to Claude 2.0

Claude V3 Sonnet

For the vast majority of workloads, Sonnet is faster on inputs and outputs than Anthropics Claude 2 and 2.1 models, with higher levels of intelligence. Sonnet is also more steerable, delivering more predictable and higher quality outcomes.

Claude V3.5 Sonnet

Claude 3.5 Sonnet raises the industry bar for intelligence, outperforming competitor models and Claude 3 Opus on a wide range of evaluations, with the speed and cost of our mid-tier model, Claude 3 Sonnet.

Claude V3 Haiku

The Claude 3 Haiku foundation model is the fastest and most compact model of the Claude 3 family, designed for near-instant responsiveness and seamless generative artificial intelligence experiences that mimic human interactions.

Moemate LLM

Custom fine-tuned model for role play and companion chats.

Last updated