OpenRouter 现已为所有模型提供 PDF 处理支持

Universal PDF Support

AI 导读

OpenRouter 为平台上所有模型添加了 PDF 文件处理能力，用户可在任意模型上直接读取和解析 PDF 文档。

正文 · 原文

🚀 Key Features

Works with any LLM available on OpenRouter.
Handles all types of PDFs, including image-only scans.
Automatically uses native PDF support for providers like Gemini, Anthropic, and OpenAI.
Sends embedded images from PDFs directly to multimodal models.
Can be combined with web search or Auto Router for enhanced functionality.
Accessible via API or directly from the OpenRouter Chatroom, with support for multiple PDFs and models in a single request.

Two PDF processing engines to choose from:

mistral-ocr (default): OCR support with text and embedded images extraction — $2 per 1000 pages.
pdf-text: Text-only extraction without OCR or image support — free.

Check out the full docs and implementation guide:

Video demo and more info: