OpenAI – Methods settings

Estimated reading time: 8 min

What your automation will do…

A connection to your OpenAI account is needed so we can sync it with our platform.
Refer to OpenAI – Connection

You can include between 1 and 255 Methods to build complex Flows.

Chat Completions Copy Link

The Chat Completions API endpoint will generate a model response from a list of messages comprising a conversation.

Generate AI Response Copy Link

This Method lets you talk to OpenAI’s models (like GPT-4) using a conversation format—just like chatting with ChatGPT.
You send messages to the AI, and it replies in a way that continues the conversation.

How it Works:
You send a list of messages with roles, like:
• System – sets behavior or rules for the assistant.
Example: “You are a helpful assistant.”
• User – your input or question.
Example: “What’s the weather today?”
• AI assistant – Previous messages to give context.
• User: Asks a follow-up question.

The AI replies with the next message in the conversation.

For more detail, see OpenAI API reference Create chat completion

Configuration Table Copy Link

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	Generate AI Response
Connection*	Select your connection or create one.

MAP FIELDS

Model*:

Select a the model used to generate the response, like gpt-4o or o3.

OpenAI offers a wide range of models with different capabilities, performance characteristics, and price points. Refer to the model guide to browse and compare available models.

Messages*:

list of message objects that represent a conversation between:

The user (you or your app),
The assistant (the AI),
and optionally, a system role that gives the AI instructions.

in Wiresk, each message should have a least 1 element.

+Element +Map ↓
(Add a repeating group of fields or Map it from an array. See Field Mapping).

Element 1: Represents the first message.

Role*:

Who is speaking:

System (used as pre-prompt to give context)
User-provided prompt
Al Assistant (used to give a supposed initial response).

Text*:

what the role says

Image URL:

An image to attach to the message, compatible with GTP-4 Vision model only.

(*) required field

Response sample

[
  {
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "The capital of Germany is Berlin.",
      "refusal": null,
      "annotations": []
    },
    "logprobs": null,
    "finish_reason": "stop"
  }
]

Expand

Example of a Basic Conversation Copy Link

Set Elements (message) as following:

Element 1:

Role: “System”
Text: “You are a helpful assistant.”

Element 2:

Role: “User”
Text: “What’s the capital of France?”

Element 3:

Role: “AI Assistant”
Text: “The capital of France is Paris.”

Element 4:

Role: “User”
Text: “What about Germany?”

This tells the AI:

System: “Behave like a helpful assistant.”
User: Asked a question about France.
Assistant: Replied with “Paris.”
User: Now asks a follow-up question about Germany.

ChatGPT uses this history to generate a relevant and consistent response.

How to analyze text directly from any incoming data Copy Link

This Method can be used as a smart, flexible step in your Flow. It allows you to analyze or generate text using OpenAI’s GPT models.

You can use it in two main ways:

Analyze Input from a Trigger
When a Flow starts (e.g., via a webhook, form, or app event), you can extract text from the input and pass it into the Chat Completion step.
Analyze Data from a Previous Method
You can map the output of a previous step (e.g., a support ticket, transcription, or product description) and use the AI to interpret, summarize, or act on that text.

Mapping Data into the Complete chat Method Copy Link

You can insert dynamic data from earlier steps (Triggers or Methods) using the built-in mapping tool:

In field messages.text of the user role
Type @ directly into the text box (lightning bold function)
A popup menu will appear showing all available fields from:
- The initial Trigger
- Previous Method responses
Select the field you want to insert, and it will be dynamically replaced when the Flow runs.

Example Copy Link

In the user message text field of the Complete Chat Method, enter:

Summarize the following customer issue: @form.description

Here, @form.description refers to a field captured from the form Trigger.

After executing the Complete Chat Method in your Wiresk Flow, the assistant’s response is returned in the "content" field. This field contains the AI-generated message based on the input prompt and conversation context.

Response sample

{
  "role": "assistant",
  "content": "Summary generated by ChatGPT"
}

Expand

Next Step: Mapping the `"content"` Field Copy Link

You can now use this "content" field as input for subsequent steps in your Flow. To do this:

In any following Method or action step, click into the relevant field where you want to insert dynamic content.
Type @ in the manual input box (lightning bold function), to activate the data mapping menu.
Locate and select the "content" field from the output of the Complete Chat step.
The selected field will be inserted as a dynamic variable and evaluated during Flow execution.

Moderation Copy Link

Identify potentially harmful content in text and images.

Content Moderation Copy Link

This Method is used to analyze a piece of text or image and determine whether it violates OpenAI’s content policy. This is especially useful for:
Content filtering in user-generated platforms (e.g., forums, chats, reviews)
Pre-screening messages before they’re published or processed
Moderation workflows in customer service or community management
Risk management for brands to detect and block offensive, violent, or inappropriate content

For more detail, see OpenAI docs on Moderation.

Configuration Table:

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	Content Moderation
Connection*	Select your connection or create one.

MAP FIELDS

Input Format*:

Text

A single block of plain text input.

Use Case: Use this when you are analyzing one piece of text at a time, such as a message, comment, or user input.

Array of Texts

A list (array) of multiple text strings submitted at once.

Use Case: Ideal for batch processing or when you need to moderate several messages in one API call.

Note: The moderation model will return a response for each item in the array, making it efficient for bulk checks.

Image URL

A URL that points to an image hosted online.

Get classification information for images.

Model:

omni-moderation-latest: This model and all snapshots support more categorization options and multi-modal inputs.

(*) required field

Response sample for text moderation:

{
  “id”: “modr-1002”,
  “model”: “omni-moderation-latest”,
  “results”: [
    {
      “flagged”: false,
      “categories”: {
        “harassment”: false,
        “harassment/threatening”: false,
        “sexual”: false,
        “hate”: false,
        “hate/threatening”: false,
        “illicit”: false,
        “illicit/violent”: false,
        “self-harm/intent”: false,
        “self-harm/instructions”: false,
        “self-harm”: false,
        “sexual/minors”: false,
        “violence”: false,
        “violence/graphic”: false
      },
      “category_scores”: {
        “harassment”: 0.000546674185053768,
        “harassment/threatening”: 0.000006922183404468203,
        “sexual”: 0.000008220189478350845,
        “hate”: 0.000004757645425657629,
        “hate/threatening”: 4.289333109916098e-7,
        “illicit”: 0.000009610241549947397,
        “illicit/violent”: 0.0000024682904407607285,
        “self-harm/intent”: 0.0000036478537675800286,
        “self-harm/instructions”: 0.00021296615605463458,
        “self-harm”: 0.000009838119439583208,
        “sexual/minors”: 0.000001670142184809518,
        “violence”: 0.00000861465062380632,
        “violence/graphic”: 0.0000015689549727726036
      },
      “category_applied_input_types”: {
        “harassment”: [
          “text”
        ],
        “harassment/threatening”: [
          “text”
        ],
        “sexual”: [
          “text”
        ],
        “hate”: [
          “text”
        ],
        “hate/threatening”: [
          “text”
        ],
        “illicit”: [
          “text”
        ],
        “illicit/violent”: [
          “text”
        ],
        “self-harm/intent”: [
          “text”
        ],
        “self-harm/instructions”: [
          “text”
        ],
        “self-harm”: [
          “text”
        ],
        “sexual/minors”: [
          “text”
        ],
        “violence”: [
          “text”
        ],
        “violence/graphic”: [
          “text”
        ]
      }
    }
  ]
}

Copy

{
  "id": "modr-1002",
  "model": "omni-moderation-latest",
  "results": [
    {
      "flagged": false,
      "categories": {
        "harassment": false,
        "harassment/threatening": false,
        "sexual": false,
        "hate": false,
        "hate/threatening": false,
        "illicit": false,
        "illicit/violent": false,
        "self-harm/intent": false,
        "self-harm/instructions": false,
        "self-harm": false,
        "sexual/minors": false,
        "violence": false,
        "violence/graphic": false
      },
      "category_scores": {
        "harassment": 0.000546674185053768,
        "harassment/threatening": 0.000006922183404468203,
        "sexual": 0.000008220189478350845,
        "hate": 0.000004757645425657629,
        "hate/threatening": 4.289333109916098e-7,
        "illicit": 0.000009610241549947397,
        "illicit/violent": 0.0000024682904407607285,
        "self-harm/intent": 0.0000036478537675800286,
        "self-harm/instructions": 0.00021296615605463458,
        "self-harm": 0.000009838119439583208,
        "sexual/minors": 0.000001670142184809518,
        "violence": 0.00000861465062380632,
        "violence/graphic": 0.0000015689549727726036
      },
      "category_applied_input_types": {
        "harassment": [
          "text"
        ],
        "harassment/threatening": [
          "text"
        ],
        "sexual": [
          "text"
        ],
        "hate": [
          "text"
        ],
        "hate/threatening": [
          "text"
        ],
        "illicit": [
          "text"
        ],
        "illicit/violent": [
          "text"
        ],
        "self-harm/intent": [
          "text"
        ],
        "self-harm/instructions": [
          "text"
        ],
        "self-harm": [
          "text"
        ],
        "sexual/minors": [
          "text"
        ],
        "violence": [
          "text"
        ],
        "violence/graphic": [
          "text"
        ]
      }
    }
  ]
}

Expand

Audio Copy Link

Learn how to turn audio into text or text into audio.

Text to Speech Copy Link

This Method converts any written text into a spoken voice using OpenAI’s Text-to-Speech (TTS) API. This is useful when you need to generate audio from AI responses, form input, summaries, or any mapped text content.

For more detail, see OpenAI API reference Create Speech.

Configuration Table:

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	Text to Speech
Connection*	Select your connection or create one.

MAP FIELDS

Input*:

The text to generate audio for.

Max: 4096 characters.

Voice*:

The voice to use when generating the audio.

Supported voices are:

alloy, ash, ballad, coral, echo, fable, onyx, nova, sage, shimmer, and verse.

Previews of the voices are available in the Text to speech guide.

Response Format*:

The format to audio in.

Supported formats:

MP3: The default response format for general use cases.
Opus: For internet streaming and communication, low latency.
AAC: For digital audio compression, preferred by YouTube, Android, iOS.
FLAC: For lossless audio compression, favored by audio enthusiasts for archiving.
WAV: Uncompressed WAV audio, suitable for low-latency applications to avoid decoding overhead.
PCM: Similar to WAV but contains the raw samples in 24kHz (16-bit signed, low-endian), without the header.

Model:

One of the available TTS models

(*) required field

Response sample

{
  "name": "speech.mp3",
  "ext": "mp3",
  "content": "//PkxABgxDnIV1rAABu72SyHh..."  // truncated base64 string
}

Expand

How to Decode and Listen Copy Link

To convert the base64 output into an actual audio file:

Copy the full "content" string.
Go to a free tool like GitHub’s base64 decoder.
Paste the string and download the MP3.
You can then play or distribute the audio.

Transcription Audio Copy Link

This Method converts spoken audio into written text using OpenAI’s Whisper speech-to-text model. It’s ideal for extracting transcripts from voice notes, recorded messages, or any spoken input in supported languages.

What It Does:
• Accepts a base64-encoded audio file
• Detects the language automatically
• Transcribes the spoken content into text in the original language

For more detail, see OpenAI API reference Create Transcription.

Configuration Table:

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	Transcription Audio
Connection*	Select your connection or create one.

MAP FIELDS

Audio File (Base64 encoded)*:

Provide the Base64 encoded content.

You can map this from a previous step.

For more details, refer to Base64 Encoding documentation.

Model*:

Select the OpenAI model to use (pre-trained AI system).

File Type*:

The format of the audio file.

flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.

Prompt:

An optional text to guide the model’s style or continue a previous audio segment.

⚠️ The prompt should match the audio language.

Temperature:

The sampling temperature, between 0 and 1.

Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.

If set to 0, the model will use log probability to automatically increase the temperature until certain thresholds are hit.

(*) required field

Response sample

{
  "text": "Hello, world!"
}

Expand

Translation Audio Copy Link

This Method uses OpenAI’s Whisper model to convert spoken content from any supported language into English text.

What It Does:
• Accepts a base64-encoded audio file
• Automatically detects the spoken language
• Translates the speech into English
• Returns a clean English transcription

For more detail, see OpenAI API reference Create Translation.

Configuration Table:

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	Translation Audio
Connection*	Select your connection or create one.

MAP FIELDS

Audio File (Base64 encoded)*:

Provide the Base64 encoded content.

You can map this from a previous step.

For more details, refer to Base64 Encoding documentation.

Model*:

Select the OpenAI model to use (pre-trained AI system).

Only whisper-1 (which is powered by our open source Whisper V2 model) is currently available.

File Type*:

The format of the audio file.

flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.

Prompt:

An optional text to guide the model’s style or continue a previous audio segment.

⚠️ The prompt should be in English.

Temperature:

The sampling temperature, between 0 and 1.

Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.

If set to 0, the model will use log probability to automatically increase the temperature until certain thresholds are hit.

(*) required field

Response sample

{
  "text": "Hello, world!"
}

Expand

Image Copy Link

Given a prompt and/or an input image, the model will generate a new image. Related guide: Image generation

Generate Images Copy Link

This Method is used to create images from text descriptions. You provide a written prompt (e.g., “a robot walking through a futuristic city”), and the AI generates a new image that visually represents that description.

For more detail, see OpenAI API reference Create Image.

Configuration Table:

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	Generate Images
Connection*	Select your connection or create one.

MAP FIELDS

Prompt*:

A text description of the desired image(s).

Max 1000 characters for Model dall-e-2
Max4000 characters for Model dall-e-3

Model:

The model to use for image generation.

dall-e-2 or dall-e-3

Number of Images:

The number of images to generate. Must be between 1 and 10.

For dall-e-3, only 1 is supported.

Default: 1

Quality:

The quality of the image that will be generated.

auto (default value) will automatically select the best quality for the given model.
hd and standard are supported for dall-e-3.
standard is the only option for dall-e-2.

Response Format:

The format in which generated images with dall-e-2 and dall-e-3 are returned.

url or b64_json.

⚠️ URLs are only valid for 60 minutes after the image has been generated.

Size:

The size of the generated images.

256x256, 512x512, or 1024x1024 for dall-e-2
1024x1024, 1792x1024, or 1024x1792 for dall-e-3.

Style:

The style of the generated images.

This parameter is only supported for dall-e-3.

Must be one of:

Vivid causes the model to lean towards generating hyper-real and dramatic images.
Natural causes the model to produce more natural, less hyper-real looking images.

User ID:

A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. Learn more.

(*) required field

Response sample

[
    {
      "url": "https://image.openai.com/image/abc123xyz.png"
    }
]

Expand

AI Router Copy Link

AI Router (in development) Copy Link

This Method designed to act as a Smart AI switch that dynamically interprets and routes input data using OpenAI’s language model. It helps make real-time decisions based on AI-generated logic—ideal for workflows that require dynamic routing, contextual decisions, or intelligent branching.

⚠️ Since this Method is in development:
• Response structure might change
• Features like validation, formatting rules, or error handling might still be evolving
• Ideal for internal testing or advanced builders familiar with OpenAI prompts

Configuration Table:

Input Options (Field Mapping):

Input: Allows dynamic inputs, e.g., from a Trigger or from Step responses. Input tab>uncheck “Show recommended” to see all fields).
Default Value: Select value from a defined list or specify a fixed attribute.
Manual input: Set a custom value by using the Lightning Bold feature.

Name*	AI Router
Connection*	Select your connection or create one.

MAP FIELDS

Model*:

Select the OpenAI model (like gpt-4, gpt-3.5-turbo) that processes the routing decision.

Prompt Messages*:

This field accepts system/user messages in OpenAI Chat format to guide the AI’s logic. You can feed in data, questions, or instructions that the AI will analyze.

Context Object:

This allows you to pass additional structured input (like JSON objects) to be used in the prompt for more complex routing.

+ Add Field (Manual Field Builder)

You can manually define extra data fields. These can be of the following types:

String
Integer
Float
Boolean
Array
Object

This makes the Router very flexible, you can feed in structured business data, parsed webhook payloads, or user input to fine-tune the routing decision.

(*) required field

Test run or automate your Flow Copy Link

After setting up your Flow, you can choose to:

“Run once”, your Flow will run only a single time. You can use this function to test your Flow.
“Run Scheduler” will automate your Flows with the recurrence rule you previously defined.

For more details, refer to How to run a Flow tutorial in our Help Center.

If you are using a Webhook Trigger, the Flow will initiate automatically when a webhook is received from your connected apps. This means that the Flow is automated without a scheduler and will run until you deactivate the Flow manually. Refer to Webhook documentation in our Help Center.

Learn more about Wiresk, explore our user guide.

If you need an integration that you cannot find in Wiresk, you can make a request to our team, and we will try our best to satisfy your needs.

OpenAI – Methods settings

Chat Completions Copy Link

Generate AI Response Copy Link

Configuration Table Copy Link

Example of a Basic Conversation Copy Link

How to analyze text directly from any incoming data Copy Link

Mapping Data into the Complete chat Method Copy Link

Example Copy Link

Next Step: Mapping the "content" Field Copy Link

Moderation Copy Link

Content Moderation Copy Link

Audio Copy Link

Text to Speech Copy Link

How to Decode and Listen Copy Link

Transcription Audio Copy Link

Translation Audio Copy Link

Image Copy Link

Generate Images Copy Link

AI Router Copy Link

AI Router (in development) Copy Link

Test run or automate your Flow Copy Link

Next Step: Mapping the `"content"` Field Copy Link