xAI Grok

xAI Grok is an AI assistant built by xAI designed to chat, create images, write code, and get real-time answers from the web and X.

Data integration: Skyvia supports xAI Grok connector in the Lookup component of Data Flow and Action component of Control Flow.

Backup: Skyvia Backup does not support xAI Grok backup.

Query: Skyvia Query does not support xAI Grok.

Establishing Connection

To create a connection to xAI Grok, you need to get xAI Grok API Key.

Getting Credentials

To obtain the API Key, perform the following steps:

  1. Go to https://console.x.ai/. Sign in to you xAI Grok account and click API Settings on the left.

    Grok console main page

  2. Click Create API key.

    Grok create API key

  3. Name the API key and configure settings

    Grok new API key

  4. Copy the API key.

    Grok copy API key

Creating Connection

To connect to xAI Grok, enter your API Key.

Connection Editor window

Connector Specifics

xAI Grok connector does not contain any objects as it does not provide access to any data. It provides access to xAI Grok API that allow you to send prompts and obtain results from AI models. You can use them via different actions and stored procedures that you can call in the Execute Command action.

Stored Procedures

Skyvia supports the following stored procedures for the xAI Grok connector.

GenerateResponse

To generate a response to the provided message with an AI model, use the command:

CALL GenerateResponse(:Message, :Instructions, :Model, :PreviousResponseId, :Temperature, :TopP, :MaxOutputTokens, :MaxToolCalls, :Store, :ReasoningEffort)

Parameter Name Description
Message String. The prompt for the AI. Required parameter.
Instructions String. A system message inserted into the model's context.
Model Enum. The ID of the model to use.
PreviousResponseId String. The unique ID of the previous response to the model. Works only when Store = true
Temperature Double. The sampling temperature to use, between 0 and 2. Higher values will make the result more random, lower values - more determined.
TopP Double. It is a number from 0 to 1. It uses nucleus sampling, where the model considers the results of the tokens with top_p probability mass. The value 0.1 means that only the tokens comprising the top 10% probability mass are considered. It is not recommended to specify both Temperature and TopP parameters.
MaxOutputTokens Int32. The maximum number of output tokens that can be generated.
MaxToolCalls Int32. The maximum number of total calls to built-in tools.
Store Boolean. Determines whether to store the generated response. Use true to enable storing and false to disable.
ReasoningEffort Enum. The effort the model should put into reasoning.

The procedure returns the following result fields: Response_Result, Response_Reasoning, Response_Id, Response_Status, Response_Error, Response_Object, Response_CreatedAt, Response_CompletedAt, Response_IncompleteDetails, Response_Temperature, Response_TopP, Response_MaxOutputTokens, Response_MaxTollCalls, Response_PreviousResponseId, Response_Store, Response_Model, Response_TextFormatType, Response_TextVerbosity, Response_ReasoningEffort, Response_ReasoningSummary, Response_Truncation, Response_SafetyIdentifier, Response_ParallelToolCalls, Response_ToolChoice, Response_Tools, Response_ServiceTier, Response_UsageInputTokens, Response_UsageInputTokensDetailsCachedTokens, Response_UsageOutputTokens, Response_UsageTotalTokens, Response_UsageNumSourcesUsed, Response_UsageNumServerSideToolsUsed, Response_UsageCostInUsdTicks, Response_UsageOutputTokensDetailsReasoningTokens.

AnalyzeImage

Analyzes one or more (up to three) images with an AI model. You can refer to images in prompts using their numbers (1 to 3). To do that use the command:

CALL AnalyzeImage(:Message, :Image1_Content, :Image2_Content, :Image3_Content, :Instructions, :Model, :PreviousResponseId, :Temperature, :TopP, :MaxOutputTokens, :MaxToolCalls, :Store, :ReasoningEffort)

Parameter Name Description
Message String. The prompt for the AI. This parameter is required.
Image1_Content Binary. The first image content. This parameter is required.
Image2_Content Binary. The second image content.
Image3_Content Binary. The third image content.
Instructions String. A system message inserted into the model's context.
Model Enum. The ID of the model to use.
PreviousResponseId String. The unique ID of the previous response to the model. Works only when Store = true
Temperature Double. The sampling temperature to use, between 0 and 2. Higher values will make the result more random, lower values - more determined.
TopP Double. It is a number from 0 to 1. It uses nucleus sampling, where the model considers the results of the tokens with top_p probability mass. The value 0.1 means that only the tokens comprising the top 10% probability mass are considered. It is not recommended to specify both Temperature and TopP parameters.
MaxOutputTokens Int32. The maximum number of output tokens that can be generated.
MaxToolCalls Int32. The maximum number of total calls to built-in tools.
Store Boolean. Determines whether to store the generated response.
ReasoningEffort Enum. The effort the model should put into reasoning.

The procedure returns the following result fields: Response_Result, Response_Reasoning, Response_Id, Response_Status, Response_Error, Response_Object, Response_CreatedAt, Response_CompletedAt, Response_IncompleteDetails, Response_Temperature, Response_TopP, Response_MaxOutputTokens, Response_MaxTollCalls, Response_PreviousResponseId, Response_Store, Response_Model, Response_TextFormatType, Response_TextVerbosity, Response_ReasoningEffort, Response_ReasoningSummary, Response_Truncation, Response_SafetyIdentifier, Response_ParallelToolCalls, Response_ToolChoice, Response_Tools, Response_ServiceTier, Response_UsageInputTokens, Response_UsageInputTokensDetailsCachedTokens, Response_UsageOutputTokens, Response_UsageTotalTokens, Response_UsageNumSourcesUsed, Response_UsageNumServerSideToolsUsed, Response_UsageCostInUsdTicks, Response_UsageOutputTokensDetailsReasoningTokens.

AnalyzeImageByUrl

Analyzes one or more (up to three) images provided by URL with an AI model. You can refer to images in prompts using their numbers (1 to 3).To do that use the command:

CALL AnalyzeImageByUrl(:Message, :Image1_Url, :Image2_Url, :Image3_Url, :Instructions, :Model, :PreviousResponseId, :Temperature, :TopP, :MaxOutputTokens, :MaxToolCalls, :Store, :ReasoningEffort)

Parameter Name Description
Message String. The prompt for the AI. This parameter is required.
Image1_Url String. URL of the first image. This parameter is required.
Image2_Url String. URL of the second image.
Image3_Url String. URL of the third image.
Instructions String. A system message inserted into the model's context.
Model Enum. The ID of the model to use.
PreviousResponseId String. The unique ID of the previous response to the model. Works only when Store = true
Temperature Double. The sampling temperature to use, between 0 and 2. Higher values will make the result more random, lower values - more determined.
TopP Double. It is a number from 0 to 1. It uses nucleus sampling, where the model considers the results of the tokens with top_p probability mass. The value 0.1 means that only the tokens comprising the top 10% probability mass are considered. It is not recommended to specify both Temperature and TopP parameters.
MaxOutputTokens Int32. The maximum number of output tokens that can be generated.
MaxToolCalls Int32. The maximum number of total calls to built-in tools.
Store Boolean. Determines whether to store the generated response.
ReasoningEffort Enum. The effort the model should put into reasoning.

The procedure returns the following result fields: Response_Result, Response_Reasoning, Response_Id, Response_Status, Response_Error, Response_Object, Response_CreatedAt, Response_CompletedAt, Response_IncompleteDetails, Response_Temperature, Response_TopP, Response_MaxOutputTokens, Response_MaxTollCalls, Response_PreviousResponseId, Response_Store, Response_Model, Response_TextFormatType, Response_TextVerbosity, Response_ReasoningEffort, Response_ReasoningSummary, Response_Truncation, Response_SafetyIdentifier, Response_ParallelToolCalls, Response_ToolChoice, Response_Tools, Response_ServiceTier, Response_UsageInputTokens, Response_UsageInputTokensDetailsCachedTokens, Response_UsageOutputTokens, Response_UsageTotalTokens, Response_UsageNumSourcesUsed, Response_UsageNumServerSideToolsUsed, Response_UsageCostInUsdTicks, Response_UsageOutputTokensDetailsReasoningTokens.

You can find more details in the xAI API documentation: responses.

AnalyzeFile

Analyzes one or more (up to three) files with an AI model. You can refer to files in prompts using their names. To do that, use the command:

CALL AnalyzeFile(:Message, :File1_Content, :File1_Name, :File2_Content, :File2_Name, :File3_Content, :File3_Name, :Instructions, :Model, :PreviousResponseId, :Temperature, :TopP, :MaxOutputTokens, :MaxToolCalls, :Store, :ReasoningEffort)

Parameter Name Description
Message String. The prompt for the AI. This parameter is required.
File1_Content Binary. The first file content. This parameter is required.
File1_Name String. The name of the first file. This parameter is required.
File2_Content Binary. The second file content.
File2_Name String. The name of the second file.
File3_Content Binary. The third file content.
File3_Name String. The name of the third file.
Instructions String. A system message inserted into the model's context.
Model Enum. The ID of the model to use.
PreviousResponseId String. The unique ID of the previous response to the model. Works only when Store = true
Temperature Double. The sampling temperature to use, between 0 and 2. Higher values will make the result more random, lower values - more determined.
TopP Double. It is a number from 0 to 1. It uses nucleus sampling, where the model considers the results of the tokens with top_p probability mass. The value 0.1 means that only the tokens comprising the top 10% probability mass are considered. It is not recommended to specify both Temperature and TopP parameters.
MaxOutputTokens Int32. The maximum number of output tokens that can be generated.
MaxToolCalls Int32. The maximum number of total calls to built-in tools.
Store Boolean. Determines whether to store the generated response.
ReasoningEffort Enum. The effort the model should put into reasoning.

The procedure returns the following result fields: Response_Result, Response_Reasoning, Response_Id, Response_Status, Response_Error, Response_Object, Response_CreatedAt, Response_CompletedAt, Response_IncompleteDetails, Response_Temperature, Response_TopP, Response_MaxOutputTokens, Response_MaxTollCalls, Response_PreviousResponseId, Response_Store, Response_Model, Response_TextFormatType, Response_TextVerbosity, Response_ReasoningEffort, Response_ReasoningSummary, Response_Truncation, Response_SafetyIdentifier, Response_ParallelToolCalls, Response_ToolChoice, Response_Tools, Response_ServiceTier, Response_UsageInputTokens, Response_UsageInputTokensDetailsCachedTokens, Response_UsageOutputTokens, Response_UsageTotalTokens, Response_UsageNumSourcesUsed, Response_UsageNumServerSideToolsUsed, Response_UsageCostInUsdTicks, Response_UsageOutputTokensDetailsReasoningTokens.

GenerateImage

To generate an image based on instructions, use the command:

CALL GenerateImage(:Instructions, :Model, :AspectRatio, :Resolution)

Parameter Name Description
Instructions String. Instructions describing the desired image. This parameter is required.
Model Enum. The ID of the model to use.
AspectRatio Enum. The aspect ratio of the generated media.
Resolution Enum. The resolution of the generated media.

The procedure returns the result in the ResultImage_Content field.

EditImage

Edits one or more (up to three) images with an AI model. You can refer to images in prompts using their numbers (1 to 3). To do that, use the command:

CALL EditImage(:Instructions, :Image1_Content, :Image2_Content, :Image3_Content, :Model, :AspectRatio, :Resolution)

Parameter Name Description
Instructions String. Instructions describing the desired change. This parameter is required.
Image1_Content Binary. The first image content. This parameter is required.
Image2_Content Binary. The second image content.
Image3_Content Binary. The third image content.
Model Enum. The ID of the model to use.
AspectRatio Enum. The aspect ratio of the generated media.
Resolution Enum. The resolution of the generated media.

The procedure returns the result in the ResultImage_Content field.

EditImageByUrl

Edits one or more (up to three) images provided by URL with an AI model. You can refer to images in prompts using their numbers (1 to 3). To do that, use the command:

CALL EditImageByUrl(:Instructions, :Image1_Url, :Image2_Url, :Image3_Url, :Model, :AspectRatio, :Resolution)

Parameter Name Description
Instructions String. Instructions describing the desired change. This parameter is required.
Image1_Url String. URL of the first image. This parameter is required.
Image2_Url String. URL of the second image.
Image3_Url String. URL of the third image.
Model Enum. The ID of the model to use.
AspectRatio Enum. The aspect ratio of the generated media.
Resolution Enum. The resolution of the generated media.

The procedure returns the result in the ResultImage_Content field.

GenerateAudio

To generate audio from text, use the command:

CALL GenerateAudio(:Text, :Language, :Voice, :OutputCodec, :OutputSampleRate, :OutputMp3Bitrate)

Parameter Name Description
Text String. The input text string. This parameter is required.
Language Enum. The language of the text or audio. This parameter is required.
Voice Enum. The voice to use for audio generation.
OutputCodec Enum. The codec for the output audio.
OutputSampleRate Enum. The sample rate for the output audio.
OutputMp3Bitrate Enum. The bitrate for the output MP3 audio.

The procedure returns the result in the ResultAudio_Content field.

GenerateVideo

To generate a video with an AI model, use the command:

CALL GenerateVideo(:Instructions, :Model, :AspectRatio, :Duration, :Resolution, :Image_Content, :Image_Url)

Parameter Name Description
Instructions String. Instructions describing the desired video. This parameter is required.
Model Enum. The ID of the model to use.
AspectRatio Enum. The aspect ratio of the generated media.
Duration Int32. The duration of the video in seconds.
Resolution Enum. The resolution of the generated media.
Image_Content Binary. Image content for the video generation.
Image_Url String. Image URL for the video generation.

The procedure returns the result in the ResultVideo_Content field.

EditVideo

To edit a video with an AI model, use the command:

CALL EditVideo(:Video_Content, :Instructions, :Model)

Parameter Name Description
Video_Content Binary. The input video content. This parameter is required.
Instructions String. Instructions describing the desired change. This parameter is required.
Model Enum. The ID of the model to use.

The procedure returns the result in the ResultVideo_Content field.

EditVideoByUrl

To edit a video provided by URL with an AI model, use the command:

CALL EditVideoByUrl(:Video_Url, :Instructions, :Model)

Parameter Name Description
Video_Url String. URL of the input video. This parameter is required.
Instructions String. Instructions describing the desired change. This parameter is required.
Model Enum. The ID of the model to use.

The procedure returns the following result fields: ResultVideo_Content.

Supported Actions

Skyvia supports the standard Execute Command action for xAI Grok, as well as the following custom actions:

Analyze File
Analyze Image
Analyze Image By Url
Classify Text
Edit Image
Edit Image By Url
Edit Video
Edit Video By Url
Generate Audio
Generate Image Generate Response Generate Video
Summarize Text

On This Page