AI – Run Prompt (Text)

Power Automate Connector: Encodian – General

Overview

The 'AI – Run Prompt (Text)' Power Automate action leverages artificial intelligence (Azure Open) to interpret and respond to user-defined prompts, returning both the response and the current 'Conversation'. 

Example Flow

Please refer to the following articles showcasing the 'AI - Run Prompt' Power Automate action:

• Build a GPT Chatbot in Power Apps with Encodian Flowr

Credit Count

The credit count is determined by calculating the total number of Input and Output AI tokens used to process the request per 0.01 Encodian action. Consider the following example using the 'GPT 4.1Mini' model, rounded to a two-digit positive infinity: 

Input Credits: (Input Tokens / GPT 4.1Mini Input Token Value)/100 

Output Credits:  (Output Tokens / GPT 4.1Mini Output Token Value)/100 

Total Credits: Input Credits + Output Credits

Input Credits: (1250 / 400)/100 = 0.04 Encodian Credits

Output Credits: (2557 / 100)/100 = 0.26 Encodian Credits

Total Credits: Input Credits + Output Credits = 0.30 Encodian Credits

Please refer to the following table for the Input / Output token values per model:

Default Parameters

The default 'AI – Run Prompt (Text)' flow action parameters are detailed below:

• Model: Select the OpenAI Model 
• Prompt: The prompt for processing the 'Text' value provided
• Conversation: A JSON representation of the OpenAI conversation for the associated chat session

Advanced Parameters

The advanced 'AI – Run Prompt (Text)' flow action parameters are detailed below:

• Frequency Penalty: A value that influences the probability of generated tokens appearing based on their cumulative frequency in generated text. Positive values will make tokens less likely to appear as their frequency increases and reduce the likelihood that the model will repeat the same statements verbatim. Supported range is [-2, 2]
• Maximum Output Tokens: Set the maximum limit for the number of tokens that can be generated for output tokens
• Presence Penalty: A value that influences the probability of generated tokens appearing based on their existing presence in generated text. Positive values make tokens less likely to appear when they already exist and increase the model's likelihood of outputting new topics. Supported range is [-2, 2]
• Temperature: The sampling temperature to use that controls the apparent creativity of generated completions. Higher values will make the output more random, while lower values will make the results more focused and deterministic. It is not recommended to modify temperature and top_p for the same completion request, as the interaction of these two settings is difficult to predict. Supported range is [0, 1]

Return Parameters

The 'AI – Run Prompt (Text)' flow action returns the following data. 

Action Specific Return Values

• Message - The OpenAI response message
• Conversation - The OpenAI conversation as JSON

Standard Return Values

• OperationId - The unique ID assigned to this operation.
• HttpStatusCode - The HTTP Status code for the response.
• HttpStatusMessage - The HTTP Status message for the response.
• Errors - An array of error messages should occur in case of an error.
• Operation Status - Indicates whether the operation has been completed, has been queued, or has failed.

Data Residency

Requests sent to the 'AI – Run Prompt (Text)' action will be processed within the following Azure regions:

Error Handling - 'The translation request has been throttled.'

Potentially, at times of peak load, the following error could be generated:

"The translation request has been throttled. Please contact support@encodian.com."

This will result in delayed processing execution. For paying customers who have a 'Large' or higher subscription level, a dedicated processing endpoint can be requested via support@encodian.com, which will help to alleviate this issue.