Show HN: OLLS – LLM 输入输出的厂商中立标准
1 分•作者: gpt4o•5 个月前
我们刚刚推出了开放 LLM 规范（OLLS）——一个社区驱动的标准，统一了开发者与大型语言模型（LLM）交互的方式，涵盖了 OpenAI、Anthropic、Google 等提供商。<p>目前，每个提供商都有不同的请求/响应格式，这使得集成变得非常麻烦：<p>解析响应不一致<p>切换模型需要自定义包装器<p>错误处理和元数据差异很大<p>OLLS 定义了一个简单、可扩展的 JSON 规范，用于输入（提示、参数、元数据）和输出（内容、推理、使用情况、错误）。
可以把它想象成 LLM 的 OpenAPI——可移植、可预测且与提供商无关。<p>GitHub 仓库 - <a href="https://github.com/julurisaichandu/open-llm-specification">https://github.com/julurisaichandu/open-llm-specification</a>
示例输入/输出格式、目标和路线图
欢迎贡献者、反馈和实际应用案例！<p>让我们构建一个统一的 LLM 接口——贡献想法或参与讨论
查看原文
We have just launched the Open LLM Specification (OLLS) – a community-driven standard that unifies how developers interact with large language models (LLMs) across providers such as OpenAI, Anthropic, Google, and others.<p>Right now, every provider has different request/response formats, which makes integration painful:<p>Parsing responses is inconsistent<p>Switching models needs custom wrappers<p>Error handling and metadata vary wildly<p>OLLS defines a simple, extensible JSON spec for both inputs (prompts, parameters, metadata) and outputs (content, reasoning, usage, errors).
Think of it like OpenAPI for LLMs—portable, predictable, and provider-agnostic.<p>GitHub Repo - <a href="https://github.com/julurisaichandu/open-llm-specification">https://github.com/julurisaichandu/open-llm-specification</a>
Example input/output formats, goals, and roadmap
Looking for contributors, feedback, and real-world use cases!<p>Let’s build a unified LLM interface—contribute ideas or join the discussion