Send PDFs Directly to Model for Processing
When uploading PDFs where text has not been converted or containing images such as charts and graphs, I receive a "Error Cannot Upload PDF." Some models (e.g., Claude) can visually analyze PDFs. I would like the ability to send files directly to the model for processing rather than being pre-processed to avoid this error.
Log in to comment and vote
Comments11
Azure Chipmunk
Jan 15, 2025
I too would like this. I've reported the error as a bug to TypingMind a few times, but maybe it is a deliberate design choice. I can currently get around this by exporting each page of the pdf as an image and upload those for processing. However, this is time consuming and not a very elegant solution. I would prefer to just upload the pdf with image pages inside.
Ngoc Nguyen
Jul 15, 2025
TypingMind now support native PDF parsing for ALL models (works with image-based PDFs). OpenAI, Gemini, Claude, and all custom models with vision supports.
Apr 8, 2025
Violet Lizard
Jun 19, 2025
Esto esta repetido y tiene otros 23 votos.
https://typingmind.canny.io/feature-requests/p/send-pdfs-directly-to-model-for-processing
Violet Lizard
Jun 18, 2025
Tony Dinh Do you have any updates this feature? This has become a critical feature for us.
We’ve received multiple complaints from our users. This creates a very negative perception: many users conclude that our platform doesn't work and abandon it entirely, even when we provide technical workarounds like our own external OCR tool. For older users or those with limited tech experience — which is the majority of our user base — these extra steps are an imposible barrier.
Adding this feature would significantly improve the user experience. Is there any progress in this?
Thank you in advance,
Violet Lizard
May 30, 2025
Ngoc Nguyen Just wondering if you have any rough idea of when this feature might be released? Super excited about it!"
Rose Pudding
Apr 24, 2025
looks like all providers support this now:
Claude: https://docs.anthropic.com/en/docs/build-with-claude/pdf-support?q=pdf#process-pdfs-with-claude
Gemini: https://ai.google.dev/gemini-api/docs/document-processing
ChatGPT: https://platform.openai.com/docs/guides/pdf-files?api-mode=chat#base64-encoded-files
Openrouter: https://openrouter.ai/docs/features/images-and-pdfs
Black Lily
Apr 24, 2025
That would be a game changer
Amaranth Dew
Apr 1, 2025
Claude can do this too. Please add support for both.
Violet Lizard
Jun 19, 2025
Esto ya esta en progreso en esta URL https://typingmind.canny.io/feature-requests/p/image-based-pdf-upload
Por favor votarlo ahi a ver si lo terminan antes.