video1.0<iframe src="https://www.loom.com/embed/b6fc57a6040c41dabf4155f61bcb2df0" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/b6fc57a6040c41dabf4155f61bcb2df0-de3863d6caf4994e.gif87.109In this video, I demo the changes made to the PDF extraction route, including a new route and prompt. I show how I'm testing it with a script that takes in a PDF path and a base URL, sending it to the Vercel endpoint created for this PR. I ran tests with different providers: Gemini ran out of tokens but was partially successful, while Anthropic and OpenAI successfully extracted data from the PDF. I encourage you to check the results and provide any feedback on the implementation.