Impressed? Signup for LLMWhisperer here
đź‘‹ Hello there!
Welcome to the LLMWhisperer playground!
🛝 This Playground should let you check LLMWhisperer capabilities quickly with your own documents. We also have some sample documents available here that let us show-off LLMWhisperer capabilities. Feel free to check them out!
🛠️ We jam well with Python and Javascript! Integrate LLMWhisperer into your applications with our Python and Javascript clients in a snap.
Sign up to process 100 pages free per day!
Try a sample document or upload your own
Electronically filled PDF form
Extract electronically filled out form information including selected radiobuttons and checkboxes are a very common use case in several businesses.
Mode used: High Quality with Form Elements
Handwritten forms
Extract handwritten form information including hand-checked radiobuttons and checkboxes are still pretty common in several countries.
Mode used: High Quality with Form Elements
Mixed printed and handwritten document
This is a fairly common real-world use case that LLMWhisperer handles with ease. LLMWhisperer can easily make out the difference between printed text and handwritten text, but deal with both in a unified manner.
Mode used: High Quality
Table extraction without line reproduction
If it’s a native text PDF, just use the native text mode for extraction. It’s super fast, since it avoids computationally expensive and relatively higher latency OCR. For simple tables, there is no need to enable line reproduction in the output.
Mode used: Native Text
Table extraction with vertical line reproduction
ASCII line reproduction in the output can work wonders for LLM accuracy when extracting structured information since different rows become very clear to the LLM.
Mode used: Low Cost
Table extraction with vertical and horizontal line reproduction
ASCII line reproduction in the output can work wonders for LLM accuracy when extracting structured information since different rows become very clear to the LLM.
Mode used: Low Cost
Smartphone captured receipt
A real-world use case that can be tricky to handle. LLMWhisperer’s powerful OCR engine in combination with its Layout Preservation mode can work wonders for LLM accuracy.
Mode used: High Quality
Off-orientation scans
You can’t expect users to have a steady hand at all times! Off-orientation scans are more common than you think. LLMWhisperer’s powerful pre-processing handles this with elan.
Mode used: High Quality
ACORD forms
The popular Acord forms are standard forms that several businesses big and small have come to rely on. They have form elements and are also electronically filled out. LLMWhisperer’s extraction quality of these forms is unmatched.
Mode used: High Quality with Form Elements
Native text PDF
These are the simplest documents to process under LLMWhisperer. When extracted in Native Text mode, not only is it super fast to extract, it’s also the cheapest mode to extract using!
Mode used: Native Text
This website uses cookies to enhance the user experience.