百度Unlimited-OCR现通过vLLM支持，采用R-SWA实现整书单次解析
百度开源了Unlimited-OCR，在vLLM上跑，能一次性解析整本书，内存不涨，比DeepSeek-OCR快35%，做文档OCR的好东西。
百度Unlimited-OCR现已集成到vLLM推理框架中，基于Reference Sliding Window Attention（R-SWA）机制实现恒定KV缓存，避免内存暴涨和速度下降。
该模型能在32K上下文预算下一次性转录40+页文档，且编辑距离极低。在6K输出token场景下，推理速度比DeepSeek-OCR快35%，GPU内存和吞吐量保持恒定。
🎉 Unlimited-OCR from @Baidu_Inc now runs in vLLM. One-shot parsing of entire books with constant KV cache, powered by Reference Sliding Window Attention (R-SWA).
🧠 R-SWA keeps KV cache fixed throughout decoding — no memory blowup, no slowdown, no matter how long the output gets. 📄 Transcribe 40+ pages in a single forward pass under a 32K context budget, with remarkably low edit distance even at scale. 🪶 35% faster than DeepSeek-OCR at 6K output tokens, with fully constant TPS and GPU memory.
🔗 Recipe: https://t.co/gBAvAP7vyG 🤗 Weights: https://t.co/9hnfOt3Ab7
🙏 Thanks to the @BaiduAI_News team for the collaboration.
This report is categorized under AITOP's "Models" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Unlimited-OCR, Baidu, vLLM, OCR.
Source attribution: curated from vLLM (@vllm_project). Cross-source corroboration: 2 independent sources have covered this development (including AK), confirming industry-wide signal strength.
- AK06-28 18:37Source
