产品多源确认

OpenRouter 支持 OpenAI 增量 diff 流式输出,其他模型返回完整补丁

精选理由

做实时协作或流式内容更新的开发者,现在可以用 OpenAI 模型获得增量 diff,减少带宽和延迟,建议试试 OpenRouter 的 engine 参数。

OpenRouter 宣布支持 OpenAI 模型的增量 diff 流式输出,通过 delta 事件逐步传输差异,而其他模型则返回完整补丁。用户可通过 `engine` 参数(auto、native、openrouter)控制行为。该功能让应用接收干净的结构化补丁并应用,下一轮交互时报告结果。这提升了流式输出的效率和实时性,尤其适合需要持续更新内容的场景。

原文 · OpenRouter

OpenAI models stream the diff incrementally via delta events. All other models return the complete patch as a single output. You can control the behavior with the `engine` parameter: auto (default), native, or openrouter Your app receives a clean, structured patch to apply, and reports back on the next turn. Docs: openrouter.ai/docs/guides/fe… 💬 0 🔄 0 ❤️ 6 👀 1075 📊 1 ⚡