Hi Community,
I am trying to fetch open ai’s text response by giving image as an input. I have configured a custom API based on what I read in the community but it is giving me null as an output. Can somone please look at my blocks and correct if I am doing something wrong? I am a beginner and come from a business background. Thank you so much for all the help.
Open AI reference - https://platform.openai.com/docs/guides/vision
Below is my current configuration -
One thing you can try is to set a Text Input’s text (inside the Post function call block) to GPT Vision’s Body. Then copy and paste that into ChatGPT or Microsoft Copilot – be sure to remove or change the Authorization value! – and then ask how to fix it for Thunkable. That should point you in the right direction.
You say you’re getting a “null” value but what is the full JSON response you’re getting (from the green response block)?
Hi @tatiang - Thank you so much for the prompt revert. When I try to get full JSON reponse, it shows me blue screen. I am a beginner in this and have just started to explore Thunkable. I tried asking chatgpt but nothing helpful came from there.
I looked over your blocks again and it looks like you didn’t include the "Bearer " string (make sure to include the space after Bearer) before your API key.
If it still fails, I would focus on getting the curl command to work in a terminal or client before coming back to the Thunkable blocks. For what it’s worth, your blocks all look correct except for the Bearer string so the curl step may not be necessary.
Finally, if you still need help after that then I would recommend pasting the value from GPT Vision’s Body into a forum post here so we can see the JSON you’re sending to the API. Make sure to format it using the </> button in the toolbar that removes smart quotes that could invalidate the JSON.
Thank you so much @tatiang. I really appreciate your prompt reverts. I will try to debug it at my end and get back to you
