SpicyChat
  • πŸ‘‹Welcome to SpicyChat.AI
  • Product Guides
    • πŸ”·Profile Settings
    • πŸ‘€Characters
    • πŸ’¬Character Chats
    • 🎨Creating Characters
      • Name
      • Title
      • Greeting
      • Personality
      • Visibility
      • Avatar
      • Tags
      • Advanced Settings
    • 🎬Director Mode
    • ⭐Premium Features
      • 🎭User Personas
      • πŸ–ΌοΈConversation Images
      • πŸ”ŠText-to-Speech
      • πŸ“Semantic Memory 2.0
      • πŸ€–AI Models
      • 🐳DeepSeek V3
      • 🧠Memory Manager
  • πŸ›οΈCommunity Guidelines
  • Advanced
    • βš™οΈTokens
    • πŸ”§Generation Settings
    • Markdowns
    • πŸ–ΌοΈImage Prompts
    • πŸ“₯Importing Characters
    • πŸ› οΈOpenAI API
    • πŸ“²Mobile Apps
  • ❓FAQs
  • ❓Support
  • πŸ“šExternal Links
Powered by GitBook
On this page
  • How Tokens Affect Responses
  • How Tokens Affect Characters Creation
  • How Tokens Affect Memory
  • Getting The Most Out Of Tokens

Was this helpful?

  1. Advanced

Tokens

PreviousCommunity GuidelinesNextGeneration Settings

Last updated 23 days ago

Was this helpful?

Tokens are the individual units of text that SpicyChat processes to read your messages, portray a character, and generate responses. Every message you send, character you create, and reply you receive uses tokens.

The amount of tokens you have access to depends on the you are using and the subscription tier you subscribe to.

  • Free And Get A Taste Users: Up to 2048 Tokens.

  • True Supporter Users: Up to 4096 Tokens.

  • I’m All In Users: Up to 8192 Tokens.

How Tokens Affect Responses

Characters use tokens to respond to your messages, the more tokens they have access to, the longer their responses can be. If you go over that limit, responses will get cut off.

Every word (or piece of a word), symbol, and emoji becomes a token.

  • If the token budget is low, the character might respond with shorter, less detailed messages.

  • If a conversation gets too long, characters may forget earlier parts of the chat to stay within the token limit.

  • Characters are given a specific number of tokens per reply. If they run out of tokens, responses may get cut off.

How many tokens are available for a character to use to reply to you are dependent on your settings and your subscription tier.

  • Free and Get A Taste Users: Up to 180 Tokens.

  • True Supporter and I’m All In Users: Up to 300 Tokens.

You can adjust the amount of tokens you want characters to have access to within the Generation Settings. Learn how to access these settings and adjust the token limit .

How Tokens Affect Characters Creation

When creating a character, it’s important to keep an eye on your token budget. Every detail, like your character's personality, greeting, scenario, background information, and example dialogues, use tokens.

You can see how many tokens you are using for each aspect of your character by referring to the token count located underneath the text box.

In order to ensure that your characters have enough tokens to use for other aspects of the conversation, it’s a good idea to keep the total amount of tokens spent within 900 to 1,100 tokens. This will ensure there will be enough tokens left for immersive conversations and fun interactions!

How Tokens Affect Memory

Due to token limitations, there is a restriction on how much message history a character can remember. If the token budget is low, older messages may be forgotten. This can affect how well the character responds, especially in longer conversations.

To ensure better continuity, keep an eye on your token usage when creating characters on SpicyChat.

Getting The Most Out Of Tokens

To get the most out of your token budget, it's key to keep things as concise as possible and focus on the essential information. Here are some helpful tips to optimise token usage when creating characters:

  • Try to keep your character's personality description short, aiming to use a total of between 800 to 1,100 tokens.

  • Keep the personality and scenario sections clear and to the point, avoiding unnecessary fluff.

  • Try summarising or paraphrasing information to save tokens, while still keeping things clear and easy to understand.

How many tokens are available is dependent on the amount of Context Memory the AI model you are using has available. You can learn more about our AI models and how Context Memory works .

βš™οΈ
AI model
here
here