I have several daily emails with lots of inline images (not attachments), and would like to use GPT-vision on them.
However, I do not know what module should I use to:
- extract these inline images
- Apply GPT-Vision on these images.
Does anyone know how to solve this? Would be easier if it is attachment, but currently the sender only sends images via inline.
Hey @Acespade ,
Look I have sent the test email like below with 2 inline images
-
For demonstration purposes, Instead of gmail trigger I am just getting the email content using id. (The structure of output is same for both the nodes so no worries.)
-
Here is the demo workflow (I have pined the data so that you can see how it works) in on order to check the same on your side simply unpin the data.
So the outcome is that, still even if the images are attached as inline image, the n8n can easily download them if we turn on the download attachment option
- Then simply I split the attachements/images and pass it to open ai node one by one and rest you can see in the workflow.
Thanks for this, but are you able to achieve the same for Outlook?
Because when I turned on ādownload attachmentā option for Outlook, it does not detect any attachment for these inline images.
@Acespade , I donāt have access to the outlook account to test this at the momentā¦
But I feel it should work,
Have you tried download message attachment node of outlook?
Please let me knowā¦if it works for youā¦else I have to create an account on my side, do some testing and get back to you!
1 Like
Yes, trying download message attachment node works. Thanks
So, unlike Gmail, Outlook requires a chain of Outlook nodes to extract the inline images.
1 Like
Anyway to grab an inline image from an gmail then send it as a āslack messageā ? This is what I am struggling with atm!
You might need add an extra Slack node at the end to send the message. Specifically āSend a Messageā node.
1 Like
Cheers @Acespade
I am glad that it finally worked for you!
Please take a moment to mark this as solution so that it can help others in future : )
Happy flogramming!
1 Like