LLMs used by Dante AI

Large Language Models (LLMs) are the core technology behind your Dante AI chatbot. They enable the chatbot to understand questions, interpret context, and generate natural, helpful responses. In simple terms, LLMs are what allow your AI chatbot to communicate in a human like way.

Choosing the right LLM helps your chatbot perform better across different use cases, whether that is answering straightforward questions or handling more detailed conversations. Each LLM varies in how it balances response quality, speed, and credit usage, allowing you to select the option that best fits your needs while maintaining efficient performance.

When you deploy your AI chatbot, you can select which LLM you want it to use. This lets you control response quality and credit usage from day one.

If you want to switch models later, you can do it anytime:

1. Click Deploy AI Chatbot to open Deployments in your dashboard
2. Select the deployment you want to update
3. Switch the LLM to a different model

See below for the specific LLMs we use:

OpenAI

‍
GPT-5 - 6 credits per response
GPT-5.1 - 6 credits per response
GPT-5.2 - 9 credits per response
GPT-5 Mini - 1.5 credits per response
GPT-5 Nano - 0.2 credits per response
GPT-4.1 – 3 credits per response
GPT-4.1 Mini – 1 credit per response
GPT-4.1 Nano – 0.15 credits per response
GPT-4o 128k – 57 credits per response
GPT-4o Mini – 0.25 credits per response
GPT-4o – 4 credits per response

Anthropic (Claude)

‍
Claude 3.7 Sonnet - 10 credits per response
Claude 4 Opus - 330 credits per response
Claude 4 Sonnet -1O credits per responsel
Claude 4.1 Opus - 330 credits per response
Claude 4.5 Sonnet - 10 credits per response
Claude 4.5 Haiku - 4 Credits per response

Google

Gemini 2.0 Flash - 0.15 credits per response
Gemini 2.0 Flash Lite - 0.15 credits per response
Gemini 2.5 Flash - 0.25 credits per response
Gemini 3 Pro - 10 credits per response
Gemini 3 Flash - 2.5 credits per response

Cohere
‍
Cohere – 1 credit per response

The number of messages your chatbot can send depends on the language model (LLM) you use. For example:

If you use GPT-4o Mini, each response consumes 0.25 credits.

This means that with the Starter plan (3,000 credits), your chatbot can respond to up to 12,000 messages per month.

For more information, visit Choosing the Right LLM for Your AI Chatbot.