Create a replica

POST /v1/replicas

Organization service token

Creates a new replica.

Headers

X-API-Version string

Default value is 2025-03-25.

application/json

Body

name string Required

The name of the replica.

Maximum length is 50.
purpose string

The purpose of the replica. This field is not used for training the replica.

Maximum length is 200.
shortDescription string Required

A short description of your replica. This field is not used for training the replica.

Maximum length is 50.
greeting string Required

The first thing your replica will say when you start a conversation with them.

Maximum length is 600.
type string

The replica type. individual: A replica of yourself. character: A replica of a character: can be anything you want. brand: A replica of a business persona or organization.

Values are individual, character, or brand. Default value is character.
ownerID string Required

The replica owner ID.
private boolean

Visibility of the replica. When set to true, only the owner will be able to find the replica and chat with it.

Default value is false.
whitelistEmails array[string(email)]

Emails of users who can use the replica.

Default value is [] (empty).
slug string Required

The slug of the replica. Slugs can be used by API consumers to determine the URLs where replicas can be found.

Maximum length is 50.
tags array[string]

The tags associated with the replica. Tags help categorize replicas and make them easier to find.

Default value is [] (empty).
profileImage string(uri)

The URL of the profile image of the replica. The image will be downloaded, optimized and stored on our servers, so the URL in the response will be different. Supported formats: .jpg, .jpeg, .png, .bmp, .webp, .avif

Format should match the following pattern: https?:\/\/[-a-zA-Z0-9@:%._+~#=]{1,256}\.[a-zA-Z0-9()]+\b([-a-zA-Z0-9()@:%_+.~#?&\/=]*). Default value is https://sensay.io/assets/default-replica-profile.webp.
suggestedQuestions array[string]

Suggested questions when starting a conversation.

Default value is [] (empty).
llm object Required
Hide llm attributes Show llm attributes object
- model string
  
  The LLM model of the replica.
  
  Values are gpt-4o, claude-3-5-haiku-latest, claude-3-7-sonnet-latest, claude-4-sonnet-20250514, grok-2-latest, grok-3-beta, deepseek-chat, o3-mini, gpt-4o-mini, huggingface-eva, or huggingface-dolphin-llama. Default value is claude-3-5-haiku-latest.
- memoryMode string
  
  The memory mode of the replica.
  
  Values are prompt-caching or rag-search. Default value is rag-search.
- systemMessage string
  
  Who is your replica? How do you want it to talk, respond and act.
  
  Maximum length is 30000.
- tools array[string]
  
  The replica's tools. Tools enable agents to interact with the world. getTokenInfo: Allows replica to get token information
  
  Values are getTokenInfo, getUdaoTokenInfo, getSensayTokenInfo, getTokenInfoMEAI, answerToLife, or toolhouse. Default value is [] (empty).
voicePreviewText string

Text that can be used to generate a voice preview.

Maximum length is 400.

Responses

201 application/json

The created replica
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates if the replica was created successfully
- uuid string(uuid) Required
  
  The replica UUID
400 application/json

Bad Request
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates the status of the request
- error string Required
  
  A text representation of the error
- fingerprint string
  
  A unique identifier of the event, useful for reporting
- request_id string Required
  
  A unique identifier of the request, useful for reporting
401 application/json

Unauthorized
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates the status of the request
- error string Required
  
  A text representation of the error
- fingerprint string
  
  A unique identifier of the event, useful for reporting
- request_id string Required
  
  A unique identifier of the request, useful for reporting
404 application/json

Not Found
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates the status of the request
- error string Required
  
  A text representation of the error
- request_id string Required
  
  A unique identifier of the request, useful for reporting
409 application/json

Conflict
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates the status of the request
- error string Required
  
  A text representation of the error
- fingerprint string
  
  A unique identifier of the event, useful for reporting
- request_id string Required
  
  A unique identifier of the request, useful for reporting
415 application/json

Unsupported Media Type
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates the status of the request
- error string Required
  
  A text representation of the error
- fingerprint string
  
  A unique identifier of the event, useful for reporting
- request_id string Required
  
  A unique identifier of the request, useful for reporting
500 application/json

Internal Server Error
Hide response attributes Show response attributes object
- success boolean Required
  
  Indicates the status of the request
- error string Required
  
  A text representation of the error
- fingerprint string
  
  A unique identifier of the event, useful for reporting
- request_id string Required
  
  A unique identifier of the request, useful for reporting
- inner_exception object
  
  The inner exception
  
  Hide inner_exception attributes Show inner_exception attributes object
  
  name string Required
  
  message string Required
  
  cause string
  
  stack string

POST /v1/replicas

curl \
 --request POST 'https://api.sensay.io/v1/replicas' \
 --header "X-ORGANIZATION-SECRET: $API_KEY" \
 --header "Content-Type: application/json" \
 --header "X-API-Version: 2025-03-25" \
 --data '{"name":"John Smith","purpose":"Acts as my AI twin for answering questions about my creative work.","shortDescription":"John Smith is a 55 year old accountant from Brooklyn who loves sports and his family.","greeting":"What would you like to know?","type":"character","ownerID":"a-user-id","private":false,"whitelistEmails":["user@domain.example"],"slug":"example-replica","tags":["male","italian"],"profileImage":"https://images.invalid/photo.jpeg","suggestedQuestions":["What is the meaning of life?"],"llm":{"model":"gpt-4o","memoryMode":"rag-search","systemMessage":"Concise, knowledgeable, empathetic and cheerful.","tools":["getTokenInfo"]},"voicePreviewText":"Hi, I'm your Sensay replica! How can I assist you today?"}'

Request examples

# Headers
X-API-Version: 2025-03-25

# Payload
{
  "name": "John Smith",
  "purpose": "Acts as my AI twin for answering questions about my creative work.",
  "shortDescription": "John Smith is a 55 year old accountant from Brooklyn who loves sports and his family.",
  "greeting": "What would you like to know?",
  "type": "character",
  "ownerID": "a-user-id",
  "private": false,
  "whitelistEmails": [
    "user@domain.example"
  ],
  "slug": "example-replica",
  "tags": [
    "male",
    "italian"
  ],
  "profileImage": "https://images.invalid/photo.jpeg",
  "suggestedQuestions": [
    "What is the meaning of life?"
  ],
  "llm": {
    "model": "gpt-4o",
    "memoryMode": "rag-search",
    "systemMessage": "Concise, knowledgeable, empathetic and cheerful.",
    "tools": [
      "getTokenInfo"
    ]
  },
  "voicePreviewText": "Hi, I'm your Sensay replica! How can I assist you today?"
}

Response examples (201)

{
  "success": true,
  "uuid": "654c2874-0b43-4d84-85fc-62ba866da1cb"
}

Response examples (400)

{
  "success": true,
  "error": "string",
  "fingerprint": "14fceadd84e74ec499afe9b0f7952d6b",
  "request_id": "xyz1::reg1:reg1::ab3c4-1234567890123-0123456789ab"
}

Response examples (401)

{
  "success": true,
  "error": "string",
  "fingerprint": "14fceadd84e74ec499afe9b0f7952d6b",
  "request_id": "xyz1::reg1:reg1::ab3c4-1234567890123-0123456789ab"
}

Response examples (404)

{
  "success": true,
  "error": "string",
  "request_id": "xyz1::reg1:reg1::ab3c4-1234567890123-0123456789ab"
}

Response examples (409)

{
  "success": true,
  "error": "string",
  "fingerprint": "14fceadd84e74ec499afe9b0f7952d6b",
  "request_id": "xyz1::reg1:reg1::ab3c4-1234567890123-0123456789ab"
}

Response examples (415)

{
  "success": true,
  "error": "string",
  "fingerprint": "14fceadd84e74ec499afe9b0f7952d6b",
  "request_id": "xyz1::reg1:reg1::ab3c4-1234567890123-0123456789ab"
}

Response examples (500)

{
  "success": true,
  "error": "string",
  "fingerprint": "14fceadd84e74ec499afe9b0f7952d6b",
  "request_id": "xyz1::reg1:reg1::ab3c4-1234567890123-0123456789ab",
  "inner_exception": {
    "name": "Server overheated",
    "cause": "Request too complicated",
    "stack": "Error: Server overheated due to an unexpected situation\n    at Object.eval (eval at <anonymous>...",
    "message": "The server overheated due to an unexpected situation"
  }
}