llama.cpp 支持多种小型 OCR 模型，可在低端设备运行

AI 摘要

llama.cpp 现已支持多种小型 OCR 模型，可在低端设备（如 4GB VRAM 的 GPU 或 CPU）上运行。文章介绍了如何使用 llama-cli 和 llama-server 运行这些模型，并通过 REST API 集成到应用中。支持的模型包括 GLM-OCR、Deepseek-OCR 等，并提供了提示词格式、量化选择及减少幻觉的建议。

正文节选

llama.cpp now supports various small OCR models that can run on low-end devices. These models are small enough to run on GPU with 4GB VRAM, and some of them can even run on CPU with decent performance. In this post, I will show you how to use these OCR models with llama.cpp. At the time of writing, llama.cpp supports the following OCR models: - LightOnOCR - Qianfan-OCR - PaddleOCR-VL (note: may have degraded performance) - GLM-OCR - Deepseek-OCR - Dots.OCR - HunyuanOCR As well as other small gen

发布时间：—

抓取时间：2026-08-07 23:54

来源机构：Hugging Face