Both chatbot types work on the same technical principle. However, there are certain considerations you should make before choosing one. The following information should help you make your decision.
When is ChantalAI the right choice for you?
------------------------------------------------
- If you don't have an OpenAI API account and don't want one.
- If you don't want content restrictions (censorship).
- If perfect multilingualism isn't important to you.
- If you don't want to use the chatbot for complex tasks such as science.
- If you want perfect role-playing on any topic.
When is OpenAI the right choice for you?
----------------------------------------------
- If you already have an OpenAI API account and want to use it.
- If content restrictions (e.g., of a sexual nature) aren't important to you.
- If you need perfect multilingualism.
- If the chatbot should also handle complex tasks such as science.
- If role-playing with content restrictions is okay for you.
Overview of the most important differences
========================================
Both chatbots have the same features, with a few minor exceptions. These are described here as follows:
1. Content Restrictions:
--------------------------
OpenAI has guidelines regarding content generation. Among other things, explicit, sexual content is not permitted. Therefore, OpenAI chatbots integrate OpenAI's moderation API. This API filters content that may violate OpenAI guidelines.
ChantalAI does not have these restrictions. This makes it ideal for any type of role-playing game. However, there is also the option to activate a filter that filters out explicit content if you do not want this content in your chatbot. You can choose between the age ratings PG, Moderate, and Adult. ChantalAI chatbots automatically activate PG mode when the chatbot is used in PG regions.
2. Chat Models
----------------
OpenAI chatbots have four models to choose from (gpt-3.5-turbo, gpt-4, gpt-4-turbo, and gpt-4o). For "normal" chatbot use in Second Life, the gpt-3.5-turbo model is perfectly sufficient. The other three models were added so that you can also use the chatbot for more complex tasks outside of normal chat conversations. An example would be scientific questions or similar.
ChantalAI chatbots have two models to choose from (chantal-gpt-3 and chantal-gpt4). However, fewer models isn't really a disadvantage. ChantalAI's models have been trained for natural, interpersonal conversations and are perfectly suited for all types of chats and role-playing in Second Life. The token costs for chantal-gpt-3 are approximately one-quarter lower than for OpenAI's gpt-3.5-turbo. The chantal-gpt-4 model is roughly in the same price range as gpt-3.5-turbo and is even more perfect for in-depth chats and role-playing games.
3. Multilingualism
--------------------
The OpenAI models have been trained for multilingualism and support virtually all common languages.
The ChantalAI model chantal-gpt-3 was trained primarily in English. It can also generate text in other languages, but not with the quality and diversity of OpenAI. The chantal-gpt-4 model has been trained in multiple languages and supports many common languages. The chantal-gpt-3 model is many times cheaper than OpenAI's gpt-3.5-turbo. The chantal-gpt-4 model is roughly the same cost as gpt-3.5-turbo.
4. API Key
-----------
To operate the OpenAI chatbots, you must have an OpenAI API key.
No API key is required for ChantalAI chatbots. You can purchase tokens for these chatbots at low cost directly from the Chantal Solutions store in Second Life. The chatbot calculates token consumption in real time. As with OpenAI, the length of the input and output text, as well as the length of the system message (bio), are used for the calculation. You can view your token balance at any time in the web interface.