WOLFRAM

Wolfram Language & System Documentation Center

"xAI" (Service Connection)

Use the xAI API with the Wolfram Language

Connecting & Authenticating

ServiceConnect["xAI"] creates a connection to the xAI API. If a previously saved connection can be found, it will be used; otherwise, a new authentication request will be launched.
Use of this connection requires internet access and an xAI account.

Requests

ServiceExecute["xAI","Requests"] gives a list of all available requests.
ServiceExecute["xAI","request",params] sends a request to the xAI API, using parameters params. The following gives possible requests.
The list of available requests and their output formats may change with API updates in different versions of the Wolfram Language.

"TestConnection" — returns Success for working connection, Failure otherwise

Text

"Chat" — create a response for the given chat conversation

Parameters:
  • "Messages"(required)a list of messages in the conversation, each given as an association with "Role" and "Content" keys
    "FrequencyPenalty"Automaticpenalize tokens based on their existing frequency in the text so far (between -2 and 2)
    "LogProbs"Automaticinclude the log probabilities on the most likely tokens, as well as the chosen tokens (between 0 and 5)
    "MaxTokens"Automaticmaximum number of tokens to generate
    "Model"Automaticname of the model to use
    "N"Automaticnumber of chat completions to return
    "PresencePenalty"Automaticpenalize new tokens based on whether they appear in the text so far (between -2 and 2)
    "StopTokens"Noneup to four strings where the API will stop generating further tokens
    "Stream"Automaticreturn the result as server-sent events
    "Temperature"Automaticsampling temperature (between 0 and 2)
    "ToolChoice"Automaticwhich (if any) tool is called by the model
    "Tools"Automaticone or more LLMTool objects available to the model
    "TotalProbabilityCutof"Nonean alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with the requested probability mass
    "User"Automaticunique identifier representing the end user
    • Image

    "ImageCreate" — create a square image given a prompt

    Parameters:
  • "Prompt"(required)text description of the desired image
    "Model"Automaticname of the model to use
    "N"Automaticnumber of images to generate
    "Quality"Automaticcontrol the quality of the result; possible values include "hd"
    "User"Automaticunique identifier representing the end user
    • Model Lists

    Model Lists

    "ChatModelList" — list models available for the "Chat" request

    "ImageModelList" — list models available for the image-related requests

    Examples

    open all close all

    Basic Examples  (1)

    Create a new connection:

    Wolfram Language code: xAI = ServiceConnect["xAI"]

    Generate a response from a chat:

    Wolfram Language code: ServiceExecute[xAI, "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there"|>}}]

    Generate an Image from a prompt:

    Wolfram Language code: ServiceExecute[xAI, "ImageCreate", {"Prompt" -> "a cat in a box"}]

    Scope  (3)

    Text  (2)

    Chat  (2)

    Respond to a chat containing multiple messages:

    Wolfram Language code: ServiceExecute["xAI", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there"|>, <|"Role" -> "User", "Content" -> "I'm bored"|>}}]

    Change the sampling temperature:

    Wolfram Language code: ServiceExecute["xAI", "Chat", {"Messages" -> {<|"Role" -> "user", "Content" -> "hello there"|>}, "Temperature" -> 1}]

    Limit the number of characters returned:

    Wolfram Language code: ServiceExecute["xAI", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there. Tell me something about bears"|>}, "MaxTokens" -> 100}]

    Return multiple completions:

    Wolfram Language code: ServiceExecute["xAI", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there. Tell me something brief about bears"|>}, "N" -> 2}]

    Allow the model to use an LLMTool:

    Wolfram Language code: tool = LLMTool["countcharacter", "string", StringLength[#string]&]
    Wolfram Language code: ServiceExecute["xAI", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "count the characters in this request"|>}, "Tools" -> tool}]

    Send multimodal input:

    Wolfram Language code: ServiceExecute["xAI", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there. Tell me something about this picture"|>, <|"Role" -> "User", "Content" -> [image]|>}}]

    Send a chat request asynchronously using ServiceSubmit and collect the response using the HandlerFunctions and HandlerFunctionsKeys options:

    Wolfram Language code: res = {};ServiceSubmit["xAI", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there"|>, <|"Role" -> "User", "Content" -> "I'm bored"|>}, "Stream" -> True}, HandlerFunctionsKeys -> "BodyChunkProcessed", HandlerFunctions -> <|"BodyChunkReceived" -> Function[AppendTo[res, #]]|>]; TaskWait[%] Query["BodyChunkProcessed", All, "ContentChunk"] /@ res//Flatten

    Image  (1)

    ImageCreate  (1)

    Create an Image:

    Wolfram Language code: ServiceExecute["xAI", "ImageCreate", {"Prompt" -> "the cutest little cat"}]

    Return multiple results:

    Wolfram Language code: ServiceExecute["xAI", "ImageCreate", {"Prompt" -> "the cutest little cat", "N" -> 2}]

    Authentication  (4)

    If no connections exist, ServiceConnect will prompt a dialog where an API key can be entered:

    Wolfram Language code: ServiceConnect["xAI"]

    The API key can also be specified using the Authentication option:

    Wolfram Language code: ServiceConnect["xAI", Authentication -> "myxAIkey"]

    Use credentials stored in SystemCredential:

    Wolfram Language code: SystemCredential["key"] = "myxAIKey"; so = ServiceConnect["xAI", Authentication -> SystemCredential["key"]]

    The credentials are stored directly by the framework, since SystemCredential["key"] evaluates to a string:

    Wolfram Language code: so["Authentication"]

    Only store the SystemCredential key name rather than its value by using RuleDelayed:

    Wolfram Language code: so2 = ServiceConnect["xAI", Authentication :> SystemCredential["key"]]; so2["Authentication"] %//ReleaseHold

    Retrieve the value of the authentication credentials used in a specific service object:

    Wolfram Language code: so = ServiceConnect["xAI", Authentication -> "myxAIkey"]; so["Authentication"]

    Overwrite the authentication credentials of an existing service object:

    Wolfram Language code: so = ServiceConnect["xAI", Authentication -> "myxAIkey"]; so["Authentication"]
    Wolfram Language code: ServiceConnect[so, Authentication -> "otherKey"]; so["Authentication"]