How to Fix Missing Required Parameter Errors from the OpenAI Node in n8n

The error message from OpenAI always names the missing parameter. Enable 'Continue On Fail' on the OpenAI node and run the workflow. Check the output for the exact error. Common messages include: 'Missing required parameter: messages', 'Missing required parameter: model', 'Missing required parameter: input', or 'you must provide a model parameter'. Note the exact field name — this tells you what to fix. If the error says 'is not valid under any of the given schemas', it means a parameter is present but has the wrong type or format.

Open the OpenAI node and check every required field. For the 'Message a Model' operation, ensure: (1) the Model field has a valid model selected (gpt-4o, gpt-4o-mini, etc.), (2) the Messages section has at least one message with a Role (user or system) and Content filled in. For the 'Classify Text for Moderation' operation, ensure the Input field is not empty. For the 'Generate an Image' operation, ensure the Prompt field is filled. If any field uses an expression, click into it to verify it resolves correctly.

The most common cause of missing parameters is expressions like {{ $json.user_message }} that resolve to undefined. This happens when the upstream node does not produce a field with that exact name. To debug: (1) Open the node with the expression. (2) In the expression editor, check the 'Result' preview — if it shows 'undefined' or is empty, the expression is the problem. (3) Check the upstream node's output by running it first and examining the Output panel. (4) Ensure the field name in the expression exactly matches the upstream output, including case sensitivity.

1// Common expression issues and fixes:
2
3// WRONG: Field name mismatch
4// Expression: {{ $json.userMessage }}
5// Actual field: $json.user_message (underscore, not camelCase)
6
7// WRONG: Nested field without correct path
8// Expression: {{ $json.message }}
9// Actual structure: $json.body.message
10// Fix: {{ $json.body.message }}
11
12// WRONG: Referencing wrong node
13// Expression: {{ $json.text }}
14// This references the PREVIOUS node. To reference a specific node:
15// {{ $('Webhook').item.json.body.text }}
16
17// SAFE: With fallback value
18// {{ $json.user_message || 'No message provided' }}

Add a Code node before the OpenAI node to validate that all required data exists. This catches missing fields before they reach the API, giving you clearer error messages and the ability to provide fallback values. Set the Code node to 'Run Once for Each Item'. Check each field your OpenAI node needs and either provide defaults or throw descriptive errors.

1const item = $input.item;
2const json = item.json;
3
4// Validate required fields for chat completion
5const userMessage = json.user_message || json.text || json.body?.message || json.query;
6const systemPrompt = json.system_prompt || 'You are a helpful assistant.';
7const model = json.model || 'gpt-4o-mini';
8
9if (!userMessage || userMessage.trim() === '') {
10  throw new Error(
11    'Missing user message. Expected field: user_message, text, body.message, or query. ' +
12    'Available fields: ' + Object.keys(json).join(', ')
13  );
14}
15
16// Pass validated data to the OpenAI node
17return [{
18  json: {
19    validated_message: userMessage.trim(),
20    system_prompt: systemPrompt,
21    model: model
22  }
23}];

If you are using the HTTP Request node to call OpenAI directly (instead of the built-in OpenAI node), you must construct the messages array correctly. Each message must have a 'role' (system, user, or assistant) and 'content' (non-empty string). A common mistake is passing the messages as a string instead of an array, or omitting the content field. Use a Code node to build the payload.

1// Code node: Build OpenAI API payload
2const userMessage = $json.validated_message;
3const systemPrompt = $json.system_prompt || 'You are a helpful assistant.';
4
5const payload = {
6  model: 'gpt-4o',
7  messages: [
8    {
9      role: 'system',
10      content: systemPrompt
11    },
12    {
13      role: 'user',
14      content: userMessage
15    }
16  ],
17  max_tokens: 2048,
18  temperature: 0.7
19};
20
21// Validate the payload
22if (!payload.messages.every(m => m.role && m.content)) {
23  throw new Error('Invalid messages array: every message must have role and content');
24}
25
26return [{ json: payload }];