Issue with "Convert to file" node

[George_Tasan]

Struggling with a workflow I’m currently working on.

I’m trying to generate an image using OpenAI and an HTTP request. I’m using the following JSON body in the HTTP request body parameter:

{
“model”: “gpt-image-1”,
“prompt”: “{{ $json.output }}”,
“n”: 1,
“size”: “1536x1024”
}

I’m mapping “{{ $json.output }}” from what the previous node generated. But as it goes through the “Convert to file” node, I repeatedly receive this message:

“The value in “data[0].b64_json” is not set”

I assume there’s something missing from the previous node. Maybe I’m mapping incorrectly, not sure.

Any guidance and help is appreciated, thanks!

[darrell_tw]

gpt-image-1 return array in result

Even you sent n=1. It’s still return an array.

Need to split out first

Here is my workflow

[George_Tasan]

Was going to try what you suggested, but ran into this instead with the below error message. My openAI account is active, so not sure why this is happening. Thanks!

{ “status”: 429, “headers”: { “alt-svc”: “h3=":443"; ma=86400”, “cf-cache-status”: “DYNAMIC”, “cf-ray”: “94a9f28abd39975d-FRA”, “connection”: “keep-alive”, “content-length”: “223”, “content-type”: “application/json; charset=utf-8”, “date”: “Wed, 04 Jun 2025 19:48:23 GMT”, “server”: “cloudflare”, “set-cookie”: “__cf_bm=HwYG4pgSipoLOpx5W67aHj.SDtMe1lA2zOzAXvypabM-1749066503-1.0.1.1-zpIvDCN3ktCi3M08KtjAMat4CbPLNJC01GOqfZWg.PhqyE3qr2kjyZoVEzbEfORtGZQ0eh944Uf42_Uh3Nsxt58jpDlVfVRHoCAcEpFK1Sk; path=/; expires=Wed, 04-Jun-25 20:18:23 GMT; domain=.api.openai.com; HttpOnly; Secure; SameSite=None, _cfuvid=3fzCCrb2oBg0dujDPRMTh9YWQXhUabbdQiYHJ7JP.DE-1749066503430-0.0.1.1-604800000; path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None”, “strict-transport-security”: “max-age=31536000; includeSubDomains; preload”, “vary”: “Origin”, “x-content-type-options”: “nosniff”, “x-request-id”: “req_3ee619383d29bf88ada8ae2c37359ce3” }, “request_id”: “req_3ee619383d29bf88ada8ae2c37359ce3”, “error”: { “message”: “Your account is not active, please check your billing details on our website.”, “type”: “billing_not_active”, “param”: null, “code”: “billing_not_active” }, “code”: “billing_not_active”, “param”: null, “type”: “billing_not_active”, “lc_error_code”: “MODEL_RATE_LIMIT”, “attemptNumber”: 3, “retriesLeft”: 0 }

[darrell_tw]

Looks like you reach rate limit

OpenAI treat different level for different rate limit

Take Free tier for example.

If you are using 4o-mini

• Request per minute is 3 times
• Request per day is 200 times

So you might need to check your tier and check the rate limit of the model you are using

https://platform.openai.com/docs/guides/rate-limits

https://platform.openai.com/docs/models/gpt-4o-mini

[George_Tasan]

Thanks Darrell. I was able to get past it with your help. Now that I’m back to the main issue, I receive the following output error in the node before “Convert to file”:

“JSON parameter needs to be valid JSON”

This is my JSON body, and I’m mapping the output parameter under “Image prompt generator”:

{
“model”: “gpt-image-1”,
“prompt”: “{{ $json.output }}”,
“n”: “1”,
“size”: “1536x1024”
}

Trying to understand where I’m messing up. Maybe the output parameter is mapped incorrectly?

[darrell_tw]

That’s because your prompt is not a valid json

You might need to control the previous AI [Image prompt generator] to make sure it gave you the right json string format

without the ```json ... ```

Then you can convert it to a valid json string.

[George_Tasan]

Oh gotcha, didn’t even catch that.

Not sure why the image prompt generator is outputting it in that format. These are the prompts I’m using and I can’t see where or why this is occurring.

Content: {{json.linkedin_post}}

For the given post content, generate 1–2 short, vivid image prompts that visually capture the core idea. Each should include:

• A short OpenAI-friendly prompt (20 words or less)
• A clear visual style (flat, 3D, cinematic, sketch, vaporwave, etc.)
• A suggested format (square, portrait, wide)

Output JSON format:
{
“image_prompt”: “…”,
“style”: “…”,
“format”: “…”
}

And then this system message:

Include a visual style (e.g. isometric, flat vector, cinematic photo, watercolor sketch)

Include a suggested format (e.g. square, 16:9 wide, vertical portrait)

[Patrick_Morgan]

Have you figured out this issue? I’m getting “JSON parameters need to be valid JSON”

[darrell_tw]

I saw you use output parser. What’s inside

If I put this in output parser. Work perfectly

---

@Patrick_Morgan

Invalid json could be many reasons. You might need to share the screenshot about the content inside.

Here is another solution for other invalid json related issue.

JSON parameter needs to be valid JSON [item 1]

try { "model": "Qubico/flux1-dev", "task_type": "txt2img", "input": { "prompt": {{ $json.message.content.toJsonString() }}, "width": 540, "height": 960 } }

[jerry765]

same issue with me here

[Patrick_Morgan]

I switch Specific Body to Use Body Below, and added the json event there.