DeepSeek 以 Apache 2.0 开源推理框架 DeepInfer
Deepseek open-sources the inference engine that powers its fastest and cheapest AI model
AI 导读
DeepSeek 以 Apache License 2.0 开源了内部推理框架 DeepInfer，它是支撑 V4.1-Flash 模型大规模快速出答案的服务软件。
摘要与补充说明
该框架基于 vLLM 构建，加入了优化的计算内核、跨多芯片的分布式处理和更快的文本生成技术。DeepSeek 同时发布 deepseek-recipe，复现官方 API 的 token 处理，使自托管环境的行为与官方一致，并称未来模型的代码也将开源。公司表示目标是帮助社区降低算力成本、减少能耗，让 AI 更可负担；文章同时指出，这一发布也会给 OpenAI、Anthropic 等西方厂商带来价格压力。
正文
Deepseek has open-sourced DeepInfer, its internal inference framework, under the Apache License 2.0. Inference is how a trained AI model handles requests and generates responses. DeepInfer is the serving software that lets Deepseek's V4.1-Flash model produce answers quickly at scale.
The framework builds on vLLM with optimized compute kernels, distributed processing across multiple chips, and faster text generation techniques. Code for future Deepseek models will also ship as open source. Deepseek is also releasing deepseek-recipe, which replicates the token processing of the official Deepseek API so self-hosted setups behave the same way.
Deepseek's models are already dirt cheap. DeepInfer fills the last gap for running them even more cheaply. The goal is "to help the community lower compute costs, reduce energy use, and make AI more affordable," the company said on X. But the release also adds pricing pressure on Western providers like OpenAI and Anthropic. Given the AI race between China and the US, the move could be politically motivated too. If Western AI companies miss their growth targets, that's a risk for the US economy and beyond.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.