Overview

The OpenAI node allows you to use AI inside your Flowmage flows.

You can generate text, analyze data, transform content, or make decisions using advanced AI models like GPT-5.

This node is perfect for adding intelligence to your automations.

When to Use This Node

Use the OpenAI node when you want to:

• Generate text (emails, summaries, descriptions)

• Analyze or classify data

• Transform input (e.g. format JSON, rewrite text)

• Extract information from unstructured data

• Add decision-making logic to flows

Configuration

Model

Select the AI model you want to use.

Example:

GPT-5

? More advanced models provide better results but may use more resources.

Reasoning Effort

Controls how much “thinking” the model does.

Options:

• Low → Fast, simple tasks

• Medium → Balanced (recommended)

• High → Complex reasoning tasks

Example:

Medium

API Key (Optional)

You can provide your own OpenAI API key.

• Leave empty → uses system default

• Fill in → uses your own account

Example:

sk-...

Instructions

This defines the behavior of the AI.

Think of this as:
? “What role should the AI take?”

Example:

You are a helpful assistant that summarizes emails in a concise and professional tone.

Input

This is the data you send to the AI.

You can:

• Write static text

• Use Flowmage variables

• Combine both

Example:

Summarize this email:
@trigger.emailBody@

How It Works

1. Flow reaches the OpenAI node

2. Instructions define the AI’s behavior

3. Input is sent to the model

4. The model processes the request

5. A response is generated

6. Output becomes available for next steps

Output / Variables

The OpenAI node returns generated content.

Example usage:

@prove.output@

? prove = your Output ID

Example Use Cases

Summarizing Data

Instructions:
Summarize the following text in 3 bullet points.

Input:
@webhook.description@

Generating Emails

Instructions:
Write a professional email reply.

Input:
Customer question: @ticket.message@

Data Transformation

Instructions:
Convert the following data into clean JSON format.

Input:
@rawData@

Decision Making

Instructions:
Determine if this lead is high quality. Answer only: YES or NO.

Input:
@leadData@

Common Mistakes

❌ Vague Instructions

The AI doesn’t know what you want.

✅ Be specific about:

• Tone

• Format

• Output style

❌ Missing Context

AI responses may be incorrect or incomplete.

✅ Provide enough input data

❌ Not Using Variables

Forgetting to pass real data.

✅ Use Flowmage variables like:

@trigger.message@

❌ Overcomplicated Prompts

Too much instruction can confuse the model.

✅ Keep instructions clear and structured

Pro Tips

• Start simple, then refine your prompt

• Use structured outputs (e.g. JSON) for automation

• Combine with conditions to act on AI results

• Use low reasoning for speed, high for complex tasks

• Reuse prompts across flows for consistency

Summary

The OpenAI node brings AI-powered intelligence into your flows, allowing you to automate tasks that normally require human thinking.

From generating content to analyzing data, this node unlocks powerful new automation possibilities.