Skip to main content
POST
/
chat
/
completions
Chat Completion
curl --request POST \
  --url https://api.animusai.co/v2/chat/completions \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "messages": [
    {
      "content": "You are having a conversation with a friend.",
      "role": "system"
    },
    {
      "content": "Hey, what have you been up to?",
      "role": "user"
    }
  ],
  "temperature": 1,
  "top_p": 1,
  "n": 1,
  "max_tokens": 150,
  "stop": [
    "<|im_end|>"
  ],
  "stream": false,
  "presence_penalty": 1,
  "frequency_penalty": 1,
  "best_of": 1,
  "top_k": 40,
  "repetition_penalty": 1,
  "min_p": 0,
  "length_penalty": 1,
  "compliance": true,
  "model": "animuslabs/Vivian-llama3.1-70b-1.0-fp8",
  "reasoning": false,
  "check_image_generation": false,
  "tools": [
    {
      "type": "function",
      "function": {
        "name": "<string>",
        "description": "<string>",
        "parameters": {}
      }
    }
  ],
  "tool_choice": "none",
  "autoTurn": false
}'
{
  "id": "<string>",
  "object": "<string>",
  "created": 123,
  "model": "<string>",
  "choices": [
    {
      "index": 123,
      "message": {
        "role": "<string>",
        "content": "<string>",
        "reasoning": "<string>",
        "image_prompt": "<string>",
        "tool_calls": [
          {
            "id": "<string>",
            "type": "function",
            "function": {
              "name": "<string>",
              "arguments": "<string>"
            }
          }
        ],
        "turns": [
          "<string>"
        ],
        "next": true
      },
      "finish_reason": "<string>",
      "compliance_violations": [
        "pedophilia"
      ]
    }
  ],
  "usage": {
    "prompt_tokens": 123,
    "completion_tokens": 123,
    "total_tokens": 123
  },
  "compliance_violations": [
    "drug_use"
  ]
}

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json
​
messages
object[]
required

A chronological list of messages that compose the current conversation.

​
temperature
number
default:1

Adjusts randomness in the response generation, with lower values yielding more predictable responses.

​
top_p
number
default:1

Filters the token set to those with cumulative probability above this threshold, influencing diversity.

​
n
integer
default:1

Number of alternate responses to generate.

​
max_tokens
integer
default:150

Caps the number of tokens in the generated response. Lacks a default to allow model-specific limits.

​
stop
string[]

A set of strings which, when generated, signal the model to cease response generation.

​
stream
boolean
default:false

If set to true, the response is streamed back to the client as it's being generated.

​
presence_penalty
number
default:1

Adjusts likelihood of new words based on their existing presence in the text. Discourages repetition when positive.

​
frequency_penalty
number
default:1

Penalizes words based on their frequency in the document to encourage diversity.

​
best_of
integer
default:1

Generates several completions server-side and returns the best. The definition of "best" depends on model and settings.

​
top_k
integer
default:40

Limits consideration to the top k tokens, diversifying outputs by reducing predictability.

​
repetition_penalty
number
default:1

Modifies likelihood of repeating tokens based on their previous occurrence, counteracting model's repetition tendency.

​
min_p
number
default:0

Sets a minimum probability threshold for tokens to be considered for generation, further filtering the possible outputs.

​
length_penalty
number
default:1

Adjusts the impact of sequence length on selection, encouraging shorter or longer responses.

​
compliance
boolean
default:true

When true, verifies if the response contains any violations such as inappropriate content.

​
model
string

The model to use for generating the response.

Example:

"animuslabs/Vivian-llama3.1-70b-1.0-fp8"

​
reasoning
boolean
default:false

When true, enables reasoning/thinking content from the model. For non-streaming responses, adds a reasoning field. For streaming, thinking content appears in the stream.

​
check_image_generation
boolean
default:false

⚠️ ALPHA FEATURE: When true, the AI analyzes its response and creates an image_prompt field if an image is requested or desired. This feature is in alpha state and not recommended for production use.

​
tools
object[]

A list of tools the model may call. Currently, only functions are supported as a tool.

​
tool_choice
default:auto

Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools.

Available options:
none,
auto
​
autoTurn
boolean
default:false

When true, splits the API response into individual conversational turns returned as an array called 'turns'.

Response

200 - application/json

The generated response along with relevant metadata.

​
id
string
required

Unique identifier for the chat completion provided by the system.

​
object
string
required

Type indicator for the object, typically 'chat.completion'.

​
created
integer
required

Unix timestamp when the response was created.

​
choices
object[]
required

Contains the generated responses, along with meta-information about each.

​
model
string

Identifier for the model used in generating the response.

​
usage
object
​
compliance_violations
enum<string>[]

List of detected content violations when compliance checking is enabled.

Example:
["drug_use"]
⌘I