Megaparse is a file parser optimized for LLM Ingestion. It can parse PDFs, DOCX, PPTX in a format that is ideal for LLMs. All of that accessible from a python package, an API, or a queue.
Hi everyone,
Today I’d like to introduce you to the new Quivr project. It a simple python package, API that helps you take in documents such as PDFs, Docx, PPTx, ... and turn them into Markown
It has several new abilities:
* OCR
* Vision Models
* Table Optimization in the extraction
* Open-source
You can use it in any of your products where you need to parse file to then send them to an LLM or simply store it
Here is how to get started:
* Go to https://github.com/QuivrHQ/MegaP...
* pip install megaparse
* Have fun
Give it a try! We’d love to hear your feedback and ideas in the comments.
This is part of Supabase mega Launch Week -> https://launchweek.dev/HOME
Love it. Markdown is becoming the de-facto in AI input processing, and proper conversion to it (without having to install a million packages) will be paramount.
Megaparse sounds super useful for prepping docs for LLMs! Love the flexibility with Python, API, or queue. Does it handle complex layouts or metadata well?
Megaparse is a really interesting tool for LLM data ingestion! 🔥 How does it handle parsing complex document structures, like multi-column layouts or mixed content (text, images, tables)? Does the OCR integration maintain accuracy across different fonts and handwriting? Also, how does the API handle large-scale batch processing—are there any optimizations for speed and efficiency with extensive datasets?
Congrats on the launch @stan_girard @amine_dirhoussi @chloe_daems
Super helpful. We are working on a product that needs something similar though we have already solved the PDF parsing problem.
Quick question - do you plan to add Excel / Spreadsheet as well? This would be super helpful.
Excited to give it a try!
There's such a huge need for this. It seems like every other week I meet someone asks me about how to get structured data from a PDF with LLMs.
Awesome tool with Megaparse! 📄✨ The ability to seamlessly parse PDFs, DOCX, and PPTX for LLM ingestion is a game-changer for data extraction. I'm curious—how does Megaparse handle complex document layouts or non-standard formats? For example, if a document has lots of embedded images or custom fonts, does it still maintain accuracy in parsing? Also, what kind of customization options do you offer for different document types or use cases?
Congrats on the launch! Megaparse looks like a game-changer for parsing docs into Markdown format. What types of files do you find it works best with?
Really nice! Open source, with OCR and table optimization, perfect for LLM workflows. Congrats to the team! 🙌
Wow, this looks super handy for integrating document parsing into LLM workflows! 🚀 Love that it's open-source and includes OCR + table optimization—makes it a no-brainer for anyone working with complex document data. Can't wait to test it out! 🔥
Love the idea, it's actually exactly what we need. And open source on top of it all ... 😚
Everyone that went through the pain of parsing slides and pdf know how big a problem that solves ;) GG team!
About Megaparse [LW24] on Product Hunt
“Open-source Document Parser to Markdown with OCR/LLMs”
Megaparse [LW24] launched on Product Hunt on December 3rd, 2024 and earned 306 upvotes and 18 comments, placing #10 on the daily leaderboard. Megaparse is a file parser optimized for LLM Ingestion. It can parse PDFs, DOCX, PPTX in a format that is ideal for LLMs. All of that accessible from a python package, an API, or a queue.
Megaparse [LW24] was featured in Developer Tools (511.1k followers) and GitHub (41.2k followers) on Product Hunt. Together, these topics include over 85k products, making this a competitive space to launch in.
Who hunted Megaparse [LW24]?
Megaparse [LW24] was hunted by Stan Girard. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Megaparse [LW24] stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Stan Girard