Hardcore Guide on How to use Chub AI Sex Chat (API Key & Chat Settings)

Among the plethora of AI-based chat applications, Chub AI stands out due to its unique features. It not only supports an extensive range of text-based role-playing, including a variety of NSFW characters, but also offers a robust platform. However, its usage, whether it involves direct character interaction or creating new characters, can be challenging.

In this article, we will provide a concise overview of Chub AI, followed by a step-by-step guide for beginners on how to utilize its primary functions. Additionally, we will offer a user experience review for each feature.

What is Chub AI

Chub AI is a sophisticated platform that serves as a character hub, enabling users to engage in conversations with a diverse array of AI characters. It also allows users to create their own characters for interaction. The platform is compatible with numerous APIs such as OpenAI or Anthropic (Claude), and it includes features like chat trees, the ability to export/import chats from existing services, and more.

Chub AI provides a frontend known as Chub Venus AI, which is capable of processing NSFW (Not Safe for Work) content. It is designed to comprehend and respond effectively to specific language, thereby offering a private environment for users to communicate more freely and openly.

Initiating an AI-Based Conversation Using API Key on Chub AI

User Login

To begin, navigate to the Venus Chub AI platform and select the ‘Login’ option located at the top left corner of the page. If you are a new user, click ‘Register’ to create an account using a valid email address. Please be aware that registration is restricted to users aged 18 and above.

Character Selection

Upon successful login, you will be presented with a variety of chat characters created by other users. To refine your search, apply filters and choose from the two available options: SFW and NSFW. Additionally, Chub AI offers over 500 tags and categories for a more personalized selection. Alternatively, use the ‘Characters’ button to manually search for your preferred character.

Once you’ve identified a character, click on its image to access the character’s profile. Here, you can review the character’s description and user reviews. To initiate a chat, click the ‘Chat’ button located beneath the character’s image.

API Configuration

At this stage, the character will greet you. However, to proceed with the chat, you need to set up your API key, as indicated in the top right corner of the page. Chub AI necessitates the use of a personal API key for chatting. If you don’t possess one, you can either register for one or purchase the specific API provided by Chub AI.

To configure your API settings, click on the relevant option. By default, the system utilizes the OpenAI API, but it also supports other large language models such as Anthropic, Kobold, Ooba, OpenRouter, and PaLM 2. Chub AI also offers its proprietary model APIs, Mars and Mercury, available for a fee.

If you have an OpenAI account, log in to the OpenAI website, locate the API option in the settings, and generate a new API key. Remember to copy or note down this key. Subsequently, input this API key into the designated field on Chub AI and select the appropriate Model version. Free users should select gpt-3.5-turbo, while those with upgraded versions should choose accordingly.

Upon clicking ‘Check API Key’, the system will verify the API’s validity. Successful verification will be indicated, and you can then proceed to chat. Please note that even free OpenAI users have a limited quota for API usage. OpenAI provides a free quota of $5, after which you will need to pay for an upgrade or use other APIs, including ChubAI’s or free API keys from Anthropic (Claude model), Kobold/Ooba, and PaLM 2.

The API settings also include other options such as ‘Pre History Instruction’ and ‘Post History Instruction’. These options allow Chub AI to call your API and fine-tune the model using pre-set prompts to influence the character’s behavior. You can select from default options or customize the character’s behavior using prompts.

The ‘Impersonation Prompt’ option enables the system AI to assist in crafting responses, which will be visible in the chat interface.

Once these settings are configured, you can commence your chat with the AI chatbot.

Key Features of Chub.ai

Diverse Chatbot Characters

Chub AI boasts an extensive collection of over 500 user-generated content (UGC) character types, predominantly related to NSFW. The majority of the top-rated characters are primarily male.

Rich-Text Generation

Leveraging the power of the API you provide, Chub AI’s text generation capabilities are primarily driven by the large language model of the API, coupled with model fine-tuning.

For instance, unlike AI chatbot software like Janitor or Joyland, Chub AI offers a wealth of comprehensive text content for each character. This includes character dialogues, interactions, and detailed descriptions of psychology, actions, and scenes, all written in an engaging style. These elements can be enjoyed as a novel or experienced as a text adventure game. The character’s actions, psychology, and scene descriptions are italicized for easy differentiation from the spoken dialogue.

Advanced Contextual Understanding and Consistency

Chub AI’s API connections facilitate intelligent, context-aware responses from AI characters, enhancing the overall user experience. The chatbots maintain remarkable consistency in their personas and dialogues, thanks to models that allow users to customize their AI partners’ settings, leading to fluid and conversational interactions.

Customizable Text/Story Content

If the content or story progression doesn’t meet your expectations, you can modify it by clicking on the image on the right. The first arrow icon allows you to continue the current content.

The second image summarizes the current character settings and story dialogue, offering further modification options. However, this feature is not recommended due to its complexity and difficulty of operation.

The third icon enables voice playback (requires a Text-to-speech API key), and the last two buttons allow you to rewrite and delete your input dialogue. Upon rewriting the story, the system automatically preserves the pre-rewrite dialogue, creating two distinct story branches for flexible story development.

By clicking the settings button in the upper right corner and selecting ‘View full chat tree’, you can explore different story versions, akin to a text adventure game.


AI-Powered Auto-Responses

Unsure how to interact with the character? The system’s AI configuration can automatically generate relevant responses based on your previous dialogue content and style.

Click on the icon next to the dialogue box, and the AI will auto-generate content. You can choose to reply directly or edit before replying.

Interestingly, the system AI appears to actively avoid generating certain content. For instance, during a conversation with a character named Sara, the AI consistently maintained a distance, contrary to the expected interaction.

Modifying Generation Settings

If the character-generated content doesn’t meet your expectations, consider adjusting the Generation Settings. Click on the upper right corner of the chat interface and select ‘Generation Setting’.

In the settings, you can adjust the ‘Temperature’, ‘Repetition Penalty’, and ‘Frequency Penalty’ for the generated content.

A lower temperature setting yields more predictable and focused responses, while higher settings produce more unpredictable and creative responses.

The repetition penalty reduces the likelihood of generating tokens that have already appeared in the text, encouraging the model to introduce new topics.

Efficiently Modifying Chat Settings

For those unfamiliar with adjusting the fundamental settings of character chat, or those who find customization overly complex, Lorebooks offer a convenient solution.

A Lorebook is a collection of keywords that, when utilized, trigger specific content to be incorporated into the prompt. This enables you to supply the AI with details about the character’s background, setting, environment, etc., without the need to include it in the character definitions that consume permanent token space.

A keyword is activated when it is used in a message. For instance, if you have a keyword in your Lorebook titled “Sex”, and the content linked to this keyword is “{{char}} is extremely passionate about Sex and would go to great lengths for it.” When you use the word “Sex” in your message, this content is added to the prompt, assisting the AI in understanding how to respond when “Sex” is mentioned.

To utilize a Lorebook, click on the icon at the top of the page, locate the Lorebook of your preference, and click to access it. Lorebooks can be about characters, scenes, or other topics.

You can also explore related information about the Lorebook and even export its JSON file. However, the crucial step is to copy the URL from the address bar, such as: “lorebooks/Kane_Richman/hentai-anime-universe-ahss.”

Next, navigate to the chat settings and enable the user Lorebook option.


Paste the copied URL into the provided field and click “Add Lorebook”. If the conditions are met, a confirmation message will indicate that the Lorebook has been successfully added.

Lorebook has two primary settings:

  • Scan Depth – This determines the number of messages to be scanned for keyword matches. A depth of 2 implies that the system will only scan the last two messages (one from you and one from the character).

  • Token Budget – This is the maximum number of tokens that Lorebook information can occupy. Once this limit is reached, subsequent matching keywords will be disregarded.

Customizing a New Character

To create a personalized character, navigate to the ‘Create Character’ button located at the top of the user interface.

The interface is bifurcated into two sections: character information and character definition. The former allows you to assign a name, upload an avatar, input tags, set visibility to public, and decide on NSFW content support.

The latter, character definition, requires significant time and effort, particularly when defining the ‘Description (Personality)’ and ‘Initial Message (Greeting)’.

The ‘Personality’ section is dedicated to outlining your character’s traits, appearance, backstory, and behaviors. This section is flexible and can include any relevant information. Plain text is the suggested format for this section.

For instance, consider the character Reiko from Chub AI. This example demonstrates how prompts can be utilized to shape various facets of a character, such as occupation, appearance, personality, preferences, dislikes, and communication style.



Prompts are instrumental in teaching the large language model about the character’s personality and behavior. To overcome limitations, prompts should be concise and direct.

There are several methods for writing bots, including W++, JSON, and plain text. JSON format, although slightly more complex, is the most efficient for bot parsing.

In the context of prompts, {{char}} is used to denote the character, while {{user}} refers to the user.

The ‘Initial Message’ is the first statement your character will make at the beginning of a conversation. The length of this message is flexible, but it should align with your objectives as it can influence the character’s subsequent behavior.

For example, consider the initial message from Reiko:

Advantages and Disadvantages

Positive Aspects of Chub AI Chat:

  • Offers over 500 user-generated content character types, predominantly not safe for work (NSFW)

  • Provides extensive text length, akin to long-form novels, with exceptional writing, vivid scenes, detailed descriptions, and insightful character psychology

  • Enables comprehensive character development, including scene setting, personality traits, potential behaviors, and diversity

  • Features AI-generated dialogue to assist users who may struggle with progressing the conversation

Areas for Enhancement:

  • The platform is not beginner-friendly, and the API operation could be more user-friendly

  • The process for character customization is overly complex, making it challenging to create a lifelike character

  • Most API functionalities are not free and require a subscription

Pricing Plan:

Chub’s pricing model is based on the usage of their APIs, named “Mercury” and “Mars”. These APIs offer extended token memory and more intricate model parameters. The “Mercury” API is priced at $5/month, while the “Mars” API costs $20/month.

Conclusion

While Chub AI excels in providing rich text content, its direct usability leaves room for improvement. It is a worthwhile option for enthusiasts of text-based adventure games. However, for users seeking a more comprehensive experience, exploring other products could be beneficial.

Contents