upload a json file and refer to it in openAI - Generate a Response Module.
What is the problem & what have you tried?
I created a single module scenario with OpenAI - Generate a Response module with a simple file prompt (see screenshot).
It is not working. Does anybody know, if i am missing some settings in the module or how to fix this? The OpenAI - Upload File module works in combination with the message assistant, but it only supports ChatGPT4.1 which does not match my needs.
Error messages or input/output bundles
I receive a response error when running the module: “[400] input: The file type you uploaded is not supported. Please try again with a pdf”
The OpenAI app Upload a file module uses the underlying OpenAI Upload file API endpoint, as part of the Assistants API. But the Assistants API is now deprecated (and so isn’t supported in GPT-5.1) and will shut down in August 2026.
The underlying OpenAI Generate a response API endpoint only supports PDF files as file inputs.
I’d recommend instead simply include JSON data in your text prompt - you can even map JSON data from the output of modules earlier in your scenario.
Thanks for the reply. That was my first attempt, but the json I use in my scenario is too big and exceeds the allowed prompt length Limit.
However, I found a workaround. If I use the OpenAI-UploadFile Module and then in my Response Module use the normal Text Prompt but add the Code Interpreter Tool and assign the file ID there I can refer to it in my prompt and chatGPT understands it.
What is your goal?
What is the problem & what have you tried?
Error messages or input/output bundles
Screenshots (scenario flow, module settings, errors)