📄 中文摘要
提供了一个无需注册的聊天界面，用户可以在此测试流行的开源模型，评估其质量、响应延迟和解码速度，以及 RAG 摘要能力。该平台允许用户设置不同的推理努力级别，以便更好地适应不同的需求。用户反馈将帮助开发团队决定是否添加更多模型，进一步丰富平台的功能。
免费尝试自托管的 LLM（GPT-OSS 20B/120B、DeepSeek、Qwen3）
出处: Try Self-Hosted LLMs for Free (GPT-OSS 20B/120B, DeepSeek, Qwen3)
发布: 2026年3月3日
📄 English Summary
Try Self-Hosted LLMs for Free (GPT-OSS 20B/120B, DeepSeek, Qwen3)
A no-signup-required chat interface is available for users to test popular open-source models, evaluating their quality, response delay, and decoding speed, as well as RAG summarization capabilities. The platform allows users to set different levels of reasoning effort to better suit various needs. User feedback will assist the development team in deciding whether to add more models, further enriching the platform's functionality.