Capricorn AI

Text generation neural network.

Best models:

Capricorn Mini

The fastest model.

Suitable for those who value response speed

Capricorn (4.4b)

Latest model.

Has functions and better text generation

All models you can use: Mini, 3.8, 4, 4.1, 4.2, 4.3, 4.3F, 4.4

---

Chat response

_GLOBAL:

POST https://api.screwltd.com/v3/ai/generate

POST https://api.screwltd.com/v3/ai/generate/:modelId

_CIS:

POST https://api.acceptancegroup.org/v3/ai/generate/

POST https://api.acceptancegroup.org/v3/ai/generate/:modelId

Makes a request to the Capricorn AI.

Headers

Name	Value
Content-Type	application/json
X-API-KEY	uuid

Body

Name	Type	Description
message*	string	Message requiring a response
version*	string	Selects model version
instructions	string	Instructions for the bot
history	array	Chat history
chatId	uuid	Chat history ID saved in the cloud.
userData	json	Information about the current user
temperature	int	Creativity of the answer, from 0 to 2.
files	byte[]	An array of files for AI analysis.
deepThinking	bool	Thinking about the answer. Available in version 4.4 and above.

Using the history attribute together with chatId is not possible. Please save the data either on our servers or on your own resources.

Local save

{
  "message": "How are you?",
  "instructions": "Use a formal communication style",
  "version": "4.4",
  "temperature": 0.7,
  "history": [
    {
      "role": "user",
      "parts": [{"text": "Hello"}]
    },
    {
      "role": "model",
      "parts": [{"text": "Hello, how can I help you?"}]
    }
  ]
}

Cloud save [Contact us]

{
  "message": "Hello!",
  "instructions": "Use a formal communication style",
  "temperature": 0.7,
  "version": "4.4",
  "chatId": "uuid"
}

All models have machine vision. You can send up to 10 pictures/audio/video for neural network analysis.

Response

200

{
  "text": "Hello my friend!",
  "duration": "0.00 ms"
}

200 (with functions)

500