Power Users/Automators 折腾/讨论/分享各种开源工具/脚本/自动化工作流👥 @zhetengsha_group📌 合集 https://t.me/zhetengsha/2🎁 恰饭 https://t.me/zhetengsha/957📢 广告投放 @xream @xream_botBuy ads: https://telega.io/c/zhetengshafeedId:55438372655431680+userId:62307599601855488http://telegram.zhetengsha.eu.orghttp://telegram.zhetengsha.eu.org/posts/6110http://telegram.zhetengsha.eu.org/posts/6110Fri, 19 Jun 2026 06:15:53 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-6110-0" popovertargetaction="show" aria-label="Open image preview: SiliconScope: 专为 Apple Silicon 打造的可视化性能监控工具感谢群友 @rainyovo233 分享• 原生 SwiftUI 界面，实时监控 CPU、GPU、ANE、Media Engine、内存带宽等核心指标，无需 sudo 权限即可使用 • 内置 AI Workload 分析视图，可识别带宽瓶颈、计算瓶颈、热限制和内存压力，帮助定位本地 LLM 推理性能问题 • 支持显示模型加载情况、GPU/CPU 分流比例、统一内存占用与交换状态，适合 LM Studio 等本地 AI 工作负载调优 "> <img src="/static/https://cdn5.telesco.pe/file/v7ViFewqp8rFT7GQU2XIXdXhLUb1UERXh3vyDTlOK6aKReW7gXpQxlLJTk-mhwhQ7bxzcya83JlMFvEiUzGyRZfVisDiXp3sPD_kc_y9B7uLGnCGjlUrusRplhdfLD_Tw4pbrsRkwJHQaAdwEq8GmW6uZjRYwoW_S-hZ40NzRGyXnwz0n4rFdRLL0TA6UxfZ1T4AMacK4KhfMXNuQRQyaxlOrwTiybHKyLdgwiUju_VN2X0M4s12NszLDc3XyqTArDP04fGthG1_HfQOz0cqp7vMQ4uBJtjLEbQ23YGo_srk_jdK8-idU9ocrm9qdBK0Nm43SSpN2M879WNI0QwZYA.jpg" alt="SiliconScope: 专为 Apple Silicon 打造的可视化性能监控工具感谢群友 @rainyovo233 分享• 原生 SwiftUI 界面，实时监控 CPU、GPU、ANE、Media Engine、内存带宽等核心指标，无需 sudo 权限即可使用 • 内置 AI Workload 分析视图，可识别带宽瓶颈、计算瓶颈、热限制和内存压力，帮助定位本地 LLM 推理性能问题 • 支持显示模型加载情况、GPU/CPU 分流比例、统一内存占用与交换状态，适合 LM Studio 等本地 AI 工作负载调优 " width="800" height="694" loading="eager" /> </button> <div class="modal" id="modal-6110-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-6110-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-6110-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>SiliconScope: 专为 Apple Silicon 打造的可视化性能监控工具</u><br /><br />感谢群友 <a href="https://t.me/rainyovo233" target="_blank" title="@rainyovo233">@rainyovo233</a> 分享<br /><br />• 原生 SwiftUI 界面，实时监控 CPU、GPU、ANE、Media Engine、内存带宽等核心指标，无需 sudo 权限即可使用 <br /><br />• 内置 AI Workload 分析视图，可识别带宽瓶颈、计算瓶颈、热限制和内存压力，帮助定位本地 <mark class="highlight">LLM</mark> 推理性能问题 <br /><br />• 支持显示模型加载情况、GPU/CPU 分流比例、统一内存占用与交换状态，适合 LM Studio 等本地 AI 工作负载调优 <br /><br /><a href="https://github.com/kennss/SiliconScope" target="_blank" rel="noopener" title="https://github.com/kennss/SiliconScope">https://github.com/kennss/SiliconScope</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327880277607583303.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/2" target="_blank" rel="noopener" title="合集">合集</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6021860481923288796.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha_group" target="_blank" rel="noopener" title="群组">群组</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327792445526380187.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/957" target="_blank" rel="noopener" title="恰饭">恰饭</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6330367862535884142.webp" alt="" loading="lazy" width="20" height="20" /><a href="https://t.me/xream_bot" target="_blank" rel="noopener" title="联系推广">联系推广</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327836885552991740.webp" alt="" loading="lazy" width="20" height="20" /> <a href="/search/result?q=%23AppleSilicon" title="#AppleSilicon">#AppleSilicon</a> <a href="/search/result?q=%23macOS" title="#macOS">#macOS</a> <a href="/search/result?q=%23%E6%80%A7%E8%83%BD%E7%9B%91%E6%8E%A7" title="#性能监控">#性能监控</a> <a href="/search/result?q=%23%E6%9C%AC%E5%9C%B0AI" title="#本地AI">#本地AI</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23SwiftUI" title="#SwiftUI">#SwiftUI</a> <a href="/search/result?q=%23%E5%BC%80%E5%8F%91%E5%B7%A5%E5%85%B7" title="#开发工具">#开发工具</a> <a href="/search/result?q=%23%E7%B3%BB%E7%BB%9F%E7%9B%91%E6%8E%A7" title="#系统监控">#系统监控</a> <a href="/search/result?q=%23ANE" title="#ANE">#ANE</a> <a href="/search/result?q=%23MediaEngine" title="#MediaEngine">#MediaEngine</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a>http://telegram.zhetengsha.eu.org/posts/5853http://telegram.zhetengsha.eu.org/posts/5853Tue, 02 Jun 2026 02:02:41 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-5853-0" popovertargetaction="show" aria-label="Open image preview: gemini-web2api: 将 Google Gemini 网页端转为 OpenAI 兼容 API • 零认证免密，支持可选 API Key 校验 • 完整 Function Calling 与流式 SSE 输出，兼容 OpenAI 与 Gemini CLI • 多模型与思考深度调节，内置联网搜索，跨平台纯 Python 实现 "> <img src="/static/https://cdn5.telesco.pe/file/tyKYbnyqymVUdew2v5ejnUSDAvk3X_aqfu2MoFdyeYWfC1jNBW4-6rvgVva411FX9lDiU51JwsKNXrfd8HpYd1jePS0HVieuwn_cjwdxdKS3BBWD2YhEtjZDQsZBVtzoYCX6fk1Eq54KH9vGqacecjLE_hLe0bwqwEIJfrI-HaW8vGchNcH6bQ25uCfZL18RJxyt_Kt5CzG6bIpahsv-bJ6mrZdEvek1z3W9JtIpxZHof-8EDgpF_5MbWkk_wLEW3HUckwsMITAvY7NmX2pXWnol5NXh8rNRVU3hXLap-KZdCyOIZj12C-CQlQGknlz-C3P8zI4R9RQ_2mbkDpL8_w.jpg" alt="gemini-web2api: 将 Google Gemini 网页端转为 OpenAI 兼容 API • 零认证免密，支持可选 API Key 校验 • 完整 Function Calling 与流式 SSE 输出，兼容 OpenAI 与 Gemini CLI • 多模型与思考深度调节，内置联网搜索，跨平台纯 Python 实现 " width="800" height="800" loading="eager" /> </button> <div class="modal" id="modal-5853-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-5853-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-5853-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>gemini-web2api: 将 Google Gemini 网页端转为 OpenAI 兼容 API </u><br /><br />• 零认证免密，支持可选 API Key 校验 <br /><br />• 完整 Function Calling 与流式 SSE 输出，兼容 OpenAI 与 Gemini CLI <br /><br />• 多模型与思考深度调节，内置联网搜索，跨平台纯 Python 实现 <br /><br /><a href="https://github.com/Sophomoresty/gemini-web2api" target="_blank" rel="noopener" title="https://github.com/Sophomoresty/gemini-web2api">https://github.com/Sophomoresty/gemini-web2api</a> <br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327880277607583303.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/2" target="_blank" rel="noopener" title="合集">合集</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6021860481923288796.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha_group" target="_blank" rel="noopener" title="群组">群组</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327792445526380187.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/957" target="_blank" rel="noopener" title="恰饭">恰饭</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6330367862535884142.webp" alt="" loading="lazy" width="20" height="20" /><a href="https://t.me/xream_bot" target="_blank" rel="noopener" title="联系推广">联系推广</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327836885552991740.webp" alt="" loading="lazy" width="20" height="20" /> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23OpenAI" title="#OpenAI">#OpenAI</a> <a href="/search/result?q=%23Google" title="#Google">#Google</a> <a href="/search/result?q=%23Gemini" title="#Gemini">#Gemini</a> <a href="/search/result?q=%23Python" title="#Python">#Python</a> <a href="/search/result?q=%23API" title="#API">#API</a> <a href="/search/result?q=%23ChatGPT" title="#ChatGPT">#ChatGPT</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23%E5%BC%80%E6%BA%90" title="#开源">#开源</a> <a href="/search/result?q=%23%E4%BB%A3%E7%90%86" title="#代理">#代理</a>http://telegram.zhetengsha.eu.org/posts/5551http://telegram.zhetengsha.eu.org/posts/5551Tue, 05 May 2026 00:01:46 GMT<u>talk-normal: 让 </u><mark class="highlight"><u>LLM</u></mark><u> 输出更短、更直接、更像真人表达的 system prompt</u><br /><br /><s>这是第几个说人话项目了来着...</s><br /><br />• 一段通用 system prompt 即可优化 GPT、Gemini、LLaMA 等模型的回答风格，显著压缩篇幅并保留关键信息<br /><br />• 内置一套可实战迭代的表达规则，覆盖客套话清理、结论前置、条件式菜单收敛等高频输出问题<br /><br />• 提供 OpenClaw、ChatGPT 自定义指令、API 等多种接入方式，安装简单，适合个人工作流和 AI 应用集成<br /><br /><a href="https://github.com/hexiecs/talk-normal" target="_blank" rel="noopener" title="https://github.com/hexiecs/talk-normal">https://github.com/hexiecs/talk-normal</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327880277607583303.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/2" target="_blank" rel="noopener" title="合集">合集</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6021860481923288796.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha_group" target="_blank" rel="noopener" title="群组">群组</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327792445526380187.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/957" target="_blank" rel="noopener" title="恰饭">恰饭</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6330367862535884142.webp" alt="" loading="lazy" width="20" height="20" /><a href="https://t.me/xream_bot" target="_blank" rel="noopener" title="联系推广">联系推广</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327836885552991740.webp" alt="" loading="lazy" width="20" height="20" /> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23PromptEngineering" title="#PromptEngineering">#PromptEngineering</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23SystemPrompt" title="#SystemPrompt">#SystemPrompt</a> <a href="/search/result?q=%23%E5%BC%80%E5%8F%91%E5%B7%A5%E5%85%B7" title="#开发工具">#开发工具</a> <a href="/search/result?q=%23%E6%95%88%E7%8E%87%E5%B7%A5%E5%85%B7" title="#效率工具">#效率工具</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a><a class="tgme_widget_message_link_preview" href="https://github.com/hexiecs/talk-normal" target="_blank" rel="noopener" title="Make any LLM talk like a normal person. A system prompt that removes AI slop. - hexiecs/talk-normal"> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - hexiecs/talk-normal: Make any LLM talk like a normal person. A system prompt that removes AI slop." src="/static/https://cdn4.telesco.pe/file/BaOn0rbYejjQG7HiymHXJzSU5eBBS8M9GsRRdei5mVT_PpLnpFB3W7a_mFp5X3_0z43xiAFTJc47z1XsOnE9CVmxbfqsrZ_byWluDzfLy2YCR4SVr38cliErgPkmEHzmiH3Cl3ySZhq0No-3eD7InA1dTFWX-jTZIIhev8LCgvn0yGdORa4YC13k_u01bwlcx-YeC9xfGwre7if5R_z33Gmo_KW1AuDfHwD-XM3t5vySeUQrVuIfg4ROnU1jIHT9B2ymMEAhs7pL9pG84Ml3Tg9R6K7HYFBpiJ4EijmGjd8NviHlOfM47k-PQJYvM-fruXFgI8HTBAYrG4OZH8rdHg.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - hexiecs/talk-normal: Make any <mark class="highlight">LLM</mark> talk like a normal person. A system prompt that removes AI slop.</div> <div class="link_preview_description">Make any <mark class="highlight">LLM</mark> talk like a normal person. A system prompt that removes AI slop. - hexiecs/talk-normal</div> </a>http://telegram.zhetengsha.eu.org/posts/5495http://telegram.zhetengsha.eu.org/posts/5495Tue, 28 Apr 2026 04:01:06 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-5495-0" popovertargetaction="show" aria-label="Open image preview: ToGo Agent: 面向 LLM 的多智能体协作开源平台• 支持多个 AI Agent 在统一群聊空间中实时协作，适合处理拆解、讨论、补位式的复杂任务• 提供角色、技能、性格等自定义能力，方便快速搭建贴近真实团队分工的 Agent 体系• 内置可视化 Web 控制台与多层级团队架构管理，覆盖任务执行、消息流转和组织配置全过程"> <img src="/static/https://cdn5.telesco.pe/file/RdO3-yuNCZtyIV8yNOtGGy9j1Zp48N4xdQ20eHTQNtP-W0pGMSkHWylVyU8CtHXE1SUPaEfQHhavn7Tpw8-ZvdPF6AepU_GkGvepaoKHvFY9T5h7C6tGera5-JsqbpD5c4cq2k_VZpqzrwLY6hO3V7zdaCDltRR7WmgkpC3c1M0zmQUgHKL9PIIT4BklcmwWoJB9IRzejr1cZIh5q-vzD90z7jRdSSkHbnkmfSLXB4LZ_KW0w4KrK2LGFKTAk8qix1qUxRYFhiYBTUpZ_VXj1v_wTJNpyYPlaA_Xzq37y96MwDrTzdV2U3vUxEUfBNGXhlI7gwCPLpfh7L0cKfN2-A.jpg" alt="ToGo Agent: 面向 LLM 的多智能体协作开源平台• 支持多个 AI Agent 在统一群聊空间中实时协作，适合处理拆解、讨论、补位式的复杂任务• 提供角色、技能、性格等自定义能力，方便快速搭建贴近真实团队分工的 Agent 体系• 内置可视化 Web 控制台与多层级团队架构管理，覆盖任务执行、消息流转和组织配置全过程" width="800" height="448" loading="eager" /> </button> <div class="modal" id="modal-5495-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-5495-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-5495-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>ToGo Agent: 面向 </u><mark class="highlight"><u>LLM</u></mark><u> 的多智能体协作开源平台</u><br /><br />• 支持多个 AI Agent 在统一群聊空间中实时协作，适合处理拆解、讨论、补位式的复杂任务<br /><br />• 提供角色、技能、性格等自定义能力，方便快速搭建贴近真实团队分工的 Agent 体系<br /><br />• 内置可视化 Web 控制台与多层级团队架构管理，覆盖任务执行、消息流转和组织配置全过程<br /><br /><a href="https://github.com/alexazhou/TogoAgent" target="_blank" rel="noopener" title="https://github.com/alexazhou/TogoAgent">https://github.com/alexazhou/TogoAgent</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327880277607583303.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/2" target="_blank" rel="noopener" title="合集">合集</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6021860481923288796.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha_group" target="_blank" rel="noopener" title="群组">群组</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327792445526380187.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/957" target="_blank" rel="noopener" title="恰饭">恰饭</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6330367862535884142.webp" alt="" loading="lazy" width="20" height="20" /><a href="https://t.me/xream_bot" target="_blank" rel="noopener" title="联系推广">联系推广</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327836885552991740.webp" alt="" loading="lazy" width="20" height="20" /> <a href="/search/result?q=%23%E5%A4%9A%E6%99%BA%E8%83%BD%E4%BD%93" title="#多智能体">#多智能体</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23Agent" title="#Agent">#Agent</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23%E5%BC%80%E6%BA%90%E9%A1%B9%E7%9B%AE" title="#开源项目">#开源项目</a> <a href="/search/result?q=%23%E5%8D%8F%E4%BD%9C%E7%B3%BB%E7%BB%9F" title="#协作系统">#协作系统</a> <a href="/search/result?q=%23%E5%8F%AF%E8%A7%86%E5%8C%96%E6%8E%A7%E5%88%B6%E5%8F%B0" title="#可视化控制台">#可视化控制台</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a>http://telegram.zhetengsha.eu.org/posts/5379http://telegram.zhetengsha.eu.org/posts/5379Wed, 15 Apr 2026 10:04:38 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-5379-0" popovertargetaction="show" aria-label="Open image preview: Koharu: 基于 Rust 构建的本地优先漫画翻译工作台• 集成文本区域检测、OCR、去字修复与 LLM 翻译，覆盖漫画翻译的完整处理链路• 支持本地与远程模型后端，兼容 llama.cpp、OpenAI 兼容接口，以及 OpenAI、Gemini、Claude、DeepSeek 等服务• 提供竖排 CJK、RTL 文本渲染、可编辑 PSD 导出、Local HTTP API 与 MCP Server，方便接入自动化工作流"> <img src="/static/https://cdn5.telesco.pe/file/t8dAuHK65zlHADzooyDCiGcz8BHyl3Yih3Beut9Fe9wfpybF_jHoi-eMcxdy1a4n9WDx7YLmT2lXSDUFS6V9XIO1rIyMlOSL25gX33kGWEPxdntf5S7p6BOGGmCDIY1Xx-A_KkpWxc8Ke8Wp2If9l4ZKHrCc3uy1PJwGHS835RXqZN6yQ_b0Eg75-6u7VN2K48Q3DLgeaTR0GzmCmTQ8jStQtMv4Yis_ZMdGcUU0hP6qXwmwrFDBv27myMBqhW0gAquxDq-qdQuDtevjLWcG-tB8_iNyD427a2a5LNLNNbA8JxJF2GQwS-B4I6pt-KDCjGBov53muZVW1-gKF-te2w.jpg" alt="Koharu: 基于 Rust 构建的本地优先漫画翻译工作台• 集成文本区域检测、OCR、去字修复与 LLM 翻译，覆盖漫画翻译的完整处理链路• 支持本地与远程模型后端，兼容 llama.cpp、OpenAI 兼容接口，以及 OpenAI、Gemini、Claude、DeepSeek 等服务• 提供竖排 CJK、RTL 文本渲染、可编辑 PSD 导出、Local HTTP API 与 MCP Server，方便接入自动化工作流" width="800" height="601" loading="eager" /> </button> <div class="modal" id="modal-5379-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-5379-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-5379-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>Koharu: 基于 Rust 构建的本地优先漫画翻译工作台</u><br /><br />• 集成文本区域检测、OCR、去字修复与 <mark class="highlight">LLM</mark> 翻译，覆盖漫画翻译的完整处理链路<br /><br />• 支持本地与远程模型后端，兼容 llama.cpp、OpenAI 兼容接口，以及 OpenAI、Gemini、Claude、DeepSeek 等服务<br /><br />• 提供竖排 CJK、RTL 文本渲染、可编辑 PSD 导出、Local HTTP API 与 MCP Server，方便接入自动化工作流<br /><br /><a href="https://github.com/mayocream/koharu" target="_blank" rel="noopener" title="https://github.com/mayocream/koharu">https://github.com/mayocream/koharu</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327880277607583303.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/2" target="_blank" rel="noopener" title="合集">合集</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6021860481923288796.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha_group" target="_blank" rel="noopener" title="群组">群组</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327792445526380187.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/957" target="_blank" rel="noopener" title="恰饭">恰饭</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6330367862535884142.webp" alt="" loading="lazy" width="20" height="20" /><a href="https://t.me/xream_bot" target="_blank" rel="noopener" title="联系推广">联系推广</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327836885552991740.webp" alt="" loading="lazy" width="20" height="20" /> <a href="/search/result?q=%23%E6%BC%AB%E7%94%BB%E7%BF%BB%E8%AF%91" title="#漫画翻译">#漫画翻译</a> <a href="/search/result?q=%23OCR" title="#OCR">#OCR</a> <a href="/search/result?q=%23%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0" title="#机器学习">#机器学习</a> <a href="/search/result?q=%23Rust" title="#Rust">#Rust</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23%E6%9C%AC%E5%9C%B0%E4%BC%98%E5%85%88" title="#本地优先">#本地优先</a> <a href="/search/result?q=%23%E5%9B%BE%E5%83%8F%E5%A4%84%E7%90%86" title="#图像处理">#图像处理</a> <a href="/search/result?q=%23MCP" title="#MCP">#MCP</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a>http://telegram.zhetengsha.eu.org/posts/5282http://telegram.zhetengsha.eu.org/posts/5282Fri, 03 Apr 2026 02:04:18 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-5282-0" popovertargetaction="show" aria-label="Open image preview: Alice: 运行在 Telegram 上的常驻式 AI 社交陪伴体感谢群友 @awaEmpty 分享• 基于“压力场”模型驱动自主决策，让 AI 不再只会被动回复，而是能持续感知关系、承诺、记忆与上下文变化• 以 userbot 形态接入 Telegram，具备跨私聊、群组、频道的持续在线能力，可主动提醒、跟进对话、整理动态并维持社交联系• 支持 OpenAI 兼容接口与小模型运行，通过 LLM 直接生成 Shell Script 执行动作，兼顾实验性架构、低成本部署与可扩展技能系统"> <img src="/static/https://cdn5.telesco.pe/file/JVMG9Rt74bjI5t4URlKIUgHtXcGdj-g0Ifsz6dUpnn70IoZXIADACmtLpWpI-tn77BgLwkpXl7Y48Ylen5SZIAboDnTFaQa5fLWK1vZHVeAMq62SZ6xqEvhFIrE9D9CmzihxLnBkmF4XA0rrMKorfEWCYlpJLS2P0wQBnN65HeIOp-xacHsLoZ5FZEOX921RdxeFq2upHzdmEQtun38xWnV0jTQQZttfSUvhovvCLYt1Bz2yOu6fJJvfqShXrf_qzcd653cyvRyZCNvZrZ8cmEZd3U813APZAeY53kmVjq5wlLpbz60QSrKgGA8MJj5WoEoVJ2JMN5HmllawQkCu8A.jpg" alt="Alice: 运行在 Telegram 上的常驻式 AI 社交陪伴体感谢群友 @awaEmpty 分享• 基于“压力场”模型驱动自主决策，让 AI 不再只会被动回复，而是能持续感知关系、承诺、记忆与上下文变化• 以 userbot 形态接入 Telegram，具备跨私聊、群组、频道的持续在线能力，可主动提醒、跟进对话、整理动态并维持社交联系• 支持 OpenAI 兼容接口与小模型运行，通过 LLM 直接生成 Shell Script 执行动作，兼顾实验性架构、低成本部署与可扩展技能系统" width="800" height="393" loading="eager" /> </button> <div class="modal" id="modal-5282-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-5282-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-5282-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>Alice: 运行在 Telegram 上的常驻式 AI 社交陪伴体</u><br /><br />感谢群友 <a href="https://t.me/awaEmpty" target="_blank" title="@awaEmpty">@awaEmpty</a> 分享<br /><br />• 基于“压力场”模型驱动自主决策，让 AI 不再只会被动回复，而是能持续感知关系、承诺、记忆与上下文变化<br /><br />• 以 userbot 形态接入 Telegram，具备跨私聊、群组、频道的持续在线能力，可主动提醒、跟进对话、整理动态并维持社交联系<br /><br />• 支持 OpenAI 兼容接口与小模型运行，通过 <mark class="highlight">LLM</mark> 直接生成 Shell Script 执行动作，兼顾实验性架构、低成本部署与可扩展技能系统<br /><br /><a href="https://github.com/LlmKira/Alice" target="_blank" rel="noopener" title="https://github.com/LlmKira/Alice">https://github.com/LlmKira/Alice</a><br /><br /><a href="/search/result?q=%23Telegram" title="#Telegram">#Telegram</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23Agent" title="#Agent">#Agent</a> <a href="/search/result?q=%23%E6%95%B0%E5%AD%97%E9%99%AA%E4%BC%B4" title="#数字陪伴">#数字陪伴</a> <a href="/search/result?q=%23%E8%87%AA%E4%B8%BB%E5%86%B3%E7%AD%96" title="#自主决策">#自主决策</a> <a href="/search/result?q=%23%E7%94%A8%E6%88%B7%E6%9C%BA%E5%99%A8%E4%BA%BA" title="#用户机器人">#用户机器人</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23TypeScript" title="#TypeScript">#TypeScript</a> <a href="/search/result?q=%23%E7%A4%BE%E4%BA%A4" title="#社交">#社交</a> <a href="/search/result?q=%23Bot" title="#Bot">#Bot</a> <a href="/search/result?q=%23%E4%BA%BA%E5%BD%A2" title="#人形">#人形</a>http://telegram.zhetengsha.eu.org/posts/5280http://telegram.zhetengsha.eu.org/posts/5280Fri, 03 Apr 2026 00:03:25 GMT<u>Coe: 面向 Linux 桌面的轻量级语音输入工具</u><br /><br />• 支持热键触发录音、ASR 转写与自动回填文本，适合在 Linux 桌面应用中实现接近原生体验的语音输入<br /><br />• 提供 Fcitx5 与 GNOME 两条集成路径，可直接 CommitString 上屏，或通过 clipboard / paste 回写到当前焦点应用<br /><br />• 兼容 OpenAI、SenseVoice 与本地 Whisper.cpp，并可接入 <mark class="highlight">LLM</mark> 对转写结果做纠错清洗，兼顾灵活性与可用性<br /><br /><a href="https://github.com/quailyquaily/coe" target="_blank" rel="noopener" title="https://github.com/quailyquaily/coe">https://github.com/quailyquaily/coe</a><br /><br /><a href="/search/result?q=%23Linux" title="#Linux">#Linux</a> <a href="/search/result?q=%23%E8%AF%AD%E9%9F%B3%E8%BE%93%E5%85%A5" title="#语音输入">#语音输入</a> <a href="/search/result?q=%23%E6%A1%8C%E9%9D%A2%E5%B7%A5%E5%85%B7" title="#桌面工具">#桌面工具</a> <a href="/search/result?q=%23Fcitx5" title="#Fcitx5">#Fcitx5</a> <a href="/search/result?q=%23GNOME" title="#GNOME">#GNOME</a> <a href="/search/result?q=%23ASR" title="#ASR">#ASR</a> <a href="/search/result?q=%23Whisper" title="#Whisper">#Whisper</a>.cpp <a href="/search/result?q=%23OpenAI" title="#OpenAI">#OpenAI</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23Wayland" title="#Wayland">#Wayland</a><a class="tgme_widget_message_link_preview" href="https://github.com/quailyquaily/coe" target="_blank" rel="noopener" title="A zero-GUI Linux voice input tool. Press a hotkey, speak, press again, and the corrected text is pasted into whatever app you're using. - quailyquaily/coe"> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - quailyquaily/coe: A zero-GUI Linux voice input tool. Press a hotkey, speak, press again, and the corrected text is pasted…" src="/static/https://cdn4.telesco.pe/file/NSrrtViMNEU7dTnofdwzwT1HpTKoJ0n7N6Y1HlHl1pFXdRMG_72mKgiP9WKKCO5cBThlvAvwZj90cfdZVwHRYh-FbeED0OfSNYG_mD8u_KT3JCUAF7VGXmQnEcbST0akNblEZRE5RKGXq_BjLx2fGmI19_tIRlz7oWbZpsiSrJmsprA4c4sdaaL1p39YiWADT4RLpY53G8HbNKNi0rQPiqsRA1Pnq_oFm3BpengUdFfOmhbxXrvwWdfMDJw2f5g6dXGmZgPaNE-bwgQETeajCWiBkieMMKIMFVuwDM_GykDhqjoWbh4tP6j-rlEbjuVSDR7hOx3cYSt4H0TlWtG6eg.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - quailyquaily/coe: A zero-GUI Linux voice input tool. Press a hotkey, speak, press again, and the corrected text is pasted…</div> <div class="link_preview_description">A zero-GUI Linux voice input tool. Press a hotkey, speak, press again, and the corrected text is pasted into whatever app you're using. - quailyquaily/coe</div> </a>http://telegram.zhetengsha.eu.org/posts/5132http://telegram.zhetengsha.eu.org/posts/5132Fri, 20 Mar 2026 02:01:46 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-5132-0" popovertargetaction="show" aria-label="Open image preview: llm_benchmark: 面向复杂推理与长文本场景的 LLM 中文能力基准项目• 聚焦高难度任务设计，覆盖符号推导、规则归纳、代码阅读、日志分析、长文本总结等多类真实问题，更适合检验模型的综合推理能力• 题目设置强调指令遵循与复杂约束处理，包含生产日志、棋局解读、工具组合、寻路规划等场景，对模型实用性评估更有参考价值• 基于 GitHub 开源发布，适合用于大模型评测、能力对比、提示词测试与中文推理 benchmark 构建"> <img src="/static/https://cdn5.telesco.pe/file/tdBixwYc1GHVpVQ5b2fpa5I4MdN62yhK20ONmRd9jzlJ7SZY9h6vWiTXvYqKu_em26_fR2RfQAC_T0mFbWIwO4u_5U1_mqxpzteXVo3VrSfTKMVo5h4fIC1Qdod-FJD4OX_upwl1wjidy2cKXPNOFPqlly_zu8qN1WFSObxrucPjvjzRj_JZjR-35QWwm61wbEU_PrIklcDO5e2KcDmpZpRs_XsW-OQh0qxV03MRxtL9IscHYyIPlZREjjqh-s-wXCXDPT4eduFTY5KVngiKKl9fNXLoEpHpi2nmFbEQopK_z0Q82Nts6js7Dc0od1kvMU1KdlMv-d9VaGDJve0Wdw.jpg" alt="llm_benchmark: 面向复杂推理与长文本场景的 LLM 中文能力基准项目• 聚焦高难度任务设计，覆盖符号推导、规则归纳、代码阅读、日志分析、长文本总结等多类真实问题，更适合检验模型的综合推理能力• 题目设置强调指令遵循与复杂约束处理，包含生产日志、棋局解读、工具组合、寻路规划等场景，对模型实用性评估更有参考价值• 基于 GitHub 开源发布，适合用于大模型评测、能力对比、提示词测试与中文推理 benchmark 构建" width="800" height="626" loading="eager" /> </button> <div class="modal" id="modal-5132-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-5132-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-5132-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>llm_benchmark: 面向复杂推理与长文本场景的 </u><mark class="highlight"><u>LLM</u></mark><u> 中文能力基准项目</u><br /><br />• 聚焦高难度任务设计，覆盖符号推导、规则归纳、代码阅读、日志分析、长文本总结等多类真实问题，更适合检验模型的综合推理能力<br /><br />• 题目设置强调指令遵循与复杂约束处理，包含生产日志、棋局解读、工具组合、寻路规划等场景，对模型实用性评估更有参考价值<br /><br />• 基于 GitHub 开源发布，适合用于大模型评测、能力对比、提示词测试与中文推理 benchmark 构建<br /><br /><a href="https://github.com/llm2014/llm_benchmark" target="_blank" rel="noopener" title="https://github.com/llm2014/llm_benchmark">https://github.com/llm2014/llm_benchmark</a><br /><br /><a href="/search/result?q=%23%E5%A4%A7%E6%A8%A1%E5%9E%8B%E8%AF%84%E6%B5%8B" title="#大模型评测">#大模型评测</a> <a href="/search/result?q=%23%E4%B8%AD%E6%96%87%E5%9F%BA%E5%87%86" title="#中文基准">#中文基准</a> <a href="/search/result?q=%23%E6%8E%A8%E7%90%86%E8%83%BD%E5%8A%9B" title="#推理能力">#推理能力</a> <a href="/search/result?q=%23%E9%95%BF%E6%96%87%E6%9C%AC" title="#长文本">#长文本</a> <a href="/search/result?q=%23%E4%BB%A3%E7%A0%81%E9%98%85%E8%AF%BB" title="#代码阅读">#代码阅读</a> <a href="/search/result?q=%23%E6%97%A5%E5%BF%97%E5%88%86%E6%9E%90" title="#日志分析">#日志分析</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23Benchmark" title="#Benchmark">#Benchmark</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a>http://telegram.zhetengsha.eu.org/posts/5128http://telegram.zhetengsha.eu.org/posts/5128Thu, 19 Mar 2026 15:46:52 GMT<u>豆腐 ( Doufu ): iPhone 上的本地 HTML 项目查看与编辑工具</u><br /><br />老朋友 App 君的新作品, 简单说就是 AI 自动写 Web App<br /><br />TestFlight 亲友限定内测了一段时间 现在上架了<br /><br /><blockquote>你可以直接在设备上创建、管理和预览 HTML、CSS 和 JavaScript 项目<br /><br />每个项目运行在沙盒化的本地服务器中</blockquote><br /><br /><a href="https://apps.apple.com/app/id6760194187" target="_blank" rel="noopener" title="https://apps.apple.com/app/id6760194187">https://apps.apple.com/app/id6760194187</a><br /><br /><a href="/search/result?q=%23iOS" title="#iOS">#iOS</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23%E5%89%8D%E7%AB%AF" title="#前端">#前端</a> <a href="/search/result?q=%23Web" title="#Web">#Web</a> <a href="/search/result?q=%23App" title="#App">#App</a> <a href="/search/result?q=%23%E5%BC%80%E5%8F%91%E5%B7%A5%E5%85%B7" title="#开发工具">#开发工具</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a><a class="tgme_widget_message_link_preview" href="https://apps.apple.com/app/id6760194187" target="_blank" rel="noopener" title="Download Doufu by ZIZICICI LIMITED on the App Store. See screenshots, ratings and reviews, user tips, and more apps like Doufu."> <i class="link_preview_right_image" style="background-image:url('https://cdn4.telesco.pe/file/XSlHBoGaI8C0Vmqz7emCx8ubPPo4Bu9jgLkn53xAH6f-Q_9jtSi2FK4OawGMxHU2FHMoOwO97JpRQvcGitD_7ms6hDqxcG5yRX5HTKOBTVkQepDMoRzdcWTShxsFEOWMD1uP1X_XJ4rD8BNeR8FCahlQtOL7d4a3QcTy5K4Whf0QzO6xf0GRQg92uLio8suwZVAK1EBM4ayi0hmlLQOTeGHWE_Y_0EEG9hv8fiK9BmmgV5asPenUX6cKFCAuQ0iS_nxuXpKhILzmQ3NBrMvqUvVra6Hbjrdt3v8O1QmgW8TAPkfcEYISl0u6duts9IZTZxWnADKr5jYFxO2MrwrM1Q.jpg')"></i> <div class="link_preview_site_name accent_color">App Store</div> <div class="link_preview_title">Doufu App - App Store</div> <div class="link_preview_description">Download Doufu by ZIZICICI LIMITED on the App Store. See screenshots, ratings and reviews, user tips, and more apps like Doufu.</div> </a>http://telegram.zhetengsha.eu.org/posts/5072http://telegram.zhetengsha.eu.org/posts/5072Sat, 14 Mar 2026 02:03:54 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-5072-0" popovertargetaction="show" aria-label="Open image preview: oMLX: 面向 Mac 本地部署的高性能 LLM 推理与管理工具• 基于 Apple Silicon 优化，支持连续批处理与分层 KV Cache，可将热缓存放在内存、冷缓存落到 SSD，显著提升本地大模型的响应效率与上下文复用能力• 同一服务可统一承载 LLM、VLM、Embedding 与 Reranker，支持模型自动发现、按需加载、LRU 淘汰、Pin 常驻和 TTL 卸载，适合多模型并行工作流• 提供 macOS 菜单栏应用与 Web 管理面板，兼容 OpenAI / Anthropic API，并内置模型下载、性能测试、聊天界面与 MCP 工具集成，降低本地 AI 服务运维门槛"> <img src="/static/https://cdn5.telesco.pe/file/vDtGBoiEKrFY7d5EP8aLyJkbKyX35_9GMD-X-GwcFLHk0FV0xDPMjJDG1v3vEaCVr8TAfJKJqEw-EFJ8FyIjFw2Gbxxh8ioTbjGvKN1KzjH5nd9WOMAMWA227W0lPrYDnpbLfP7EM9mbMRosTTTpk7Ss_5l0SbS4EOIKWIHrXlFKlbfnbBh2QYdBBEKHs0VpeWFMbFs_pEDyMpgUH8xOXtNNdfcYLMeMCbEdbjkoQd7YZuDCuUkBERkpCcBHxQTWXEp6QTF9FJ21_WrTs5PYgtPBzg1DgcEP5Bi2vWO5nC48pmwmDVtOB0eEd1LD2EReEBBQJZgxEiHRYP1Yoyq6Zw.jpg" alt="oMLX: 面向 Mac 本地部署的高性能 LLM 推理与管理工具• 基于 Apple Silicon 优化，支持连续批处理与分层 KV Cache，可将热缓存放在内存、冷缓存落到 SSD，显著提升本地大模型的响应效率与上下文复用能力• 同一服务可统一承载 LLM、VLM、Embedding 与 Reranker，支持模型自动发现、按需加载、LRU 淘汰、Pin 常驻和 TTL 卸载，适合多模型并行工作流• 提供 macOS 菜单栏应用与 Web 管理面板，兼容 OpenAI / Anthropic API，并内置模型下载、性能测试、聊天界面与 MCP 工具集成，降低本地 AI 服务运维门槛" width="800" height="626" loading="eager" /> </button> <div class="modal" id="modal-5072-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-5072-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-5072-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>oMLX: 面向 Mac 本地部署的高性能 </u><mark class="highlight"><u>LLM</u></mark><u> 推理与管理工具</u><br /><br />• 基于 Apple Silicon 优化，支持连续批处理与分层 KV Cache，可将热缓存放在内存、冷缓存落到 SSD，显著提升本地大模型的响应效率与上下文复用能力<br /><br />• 同一服务可统一承载 <mark class="highlight">LLM</mark>、VLM、Embedding 与 Reranker，支持模型自动发现、按需加载、LRU 淘汰、Pin 常驻和 TTL 卸载，适合多模型并行工作流<br /><br />• 提供 macOS 菜单栏应用与 Web 管理面板，兼容 OpenAI / Anthropic API，并内置模型下载、性能测试、聊天界面与 MCP 工具集成，降低本地 AI 服务运维门槛<br /><br /><a href="https://github.com/jundot/omlx" target="_blank" rel="noopener" title="https://github.com/jundot/omlx">https://github.com/jundot/omlx</a><br /><br /><a href="/search/result?q=%23%E6%9C%AC%E5%9C%B0%E5%A4%A7%E6%A8%A1%E5%9E%8B" title="#本地大模型">#本地大模型</a> <a href="/search/result?q=%23Mac" title="#Mac">#Mac</a> AI <a href="/search/result?q=%23Apple" title="#Apple">#Apple</a> Silicon <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> 推理 <a href="/search/result?q=%23KV" title="#KV">#KV</a> Cache <a href="/search/result?q=%23%E5%A4%9A%E6%A8%A1%E5%9E%8B%E6%9C%8D%E5%8A%A1" title="#多模型服务">#多模型服务</a> <a href="/search/result?q=%23OpenAI" title="#OpenAI">#OpenAI</a> API <a href="/search/result?q=%23Anthropic" title="#Anthropic">#Anthropic</a> API <a href="/search/result?q=%23MCP" title="#MCP">#MCP</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a>http://telegram.zhetengsha.eu.org/posts/4854http://telegram.zhetengsha.eu.org/posts/4854Tue, 17 Feb 2026 07:04:15 GMT<div class="image-list-container image-list-even"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-4854-0" popovertargetaction="show" aria-label="Open image preview: Squirrel: 企业级 LLM 网关, 统一代理 OpenAI、Anthropic 及兼容 API• 一次接入统一 API, 兼容 OpenAI Chat、OpenAI Responses 与 Anthropic Messages, 并支持请求/响应/流式互转 • 智能路由与高可用, 支持规则路由、负载均衡、自动重试与故障转移, 可按成本或优先级自动选模型 • 全面可观测性 + 管理面板, 记录 Token 用量/延迟/成本与请求日志, 提供 Next.js 管理后台做供应商、模型映射与 API Key 管理 "> <img src="/static/https://cdn5.telesco.pe/file/UW9f9nVOkkEfGp-v4UYFtV_7j-wVgqNlsEt9B9lOUnIQwtjIiOAZyoYWE4WU5Svnuvpa7KpKJIfQF7l7hmc8fzWtaVoe2_DIw-KSP_pVKzCPfK2EUPcDxnL0E-Tr62gl-Meq4HyOzN_kk5Nx9vKdDvT-UA-7VqCYXz-iTlvY1UKTL02x_vXC1_80tKadIeIvyYnzdDyi3eoj5rCXQUsIcwk2nJU7f7Wlwo6KlbcZ_qpvs0Lb-deIR3DlZrQ4fCQJXjDviPD_s3kRI0hejf3ISdxohz51N3BR-AEx_1v46M-fEX1MMI1Ljdw0nJKbD-n9ZZc6P1sqz51TKhqMoopgxA.jpg" alt="Squirrel: 企业级 LLM 网关, 统一代理 OpenAI、Anthropic 及兼容 API• 一次接入统一 API, 兼容 OpenAI Chat、OpenAI Responses 与 Anthropic Messages, 并支持请求/响应/流式互转 • 智能路由与高可用, 支持规则路由、负载均衡、自动重试与故障转移, 可按成本或优先级自动选模型 • 全面可观测性 + 管理面板, 记录 Token 用量/延迟/成本与请求日志, 提供 Next.js 管理后台做供应商、模型映射与 API Key 管理 " width="225" height="357" loading="eager" /> </button> <div class="modal" id="modal-4854-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-4854-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-4854-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-4854-1" popovertargetaction="show" aria-label="Open image preview: Squirrel: 企业级 LLM 网关, 统一代理 OpenAI、Anthropic 及兼容 API• 一次接入统一 API, 兼容 OpenAI Chat、OpenAI Responses 与 Anthropic Messages, 并支持请求/响应/流式互转 • 智能路由与高可用, 支持规则路由、负载均衡、自动重试与故障转移, 可按成本或优先级自动选模型 • 全面可观测性 + 管理面板, 记录 Token 用量/延迟/成本与请求日志, 提供 Next.js 管理后台做供应商、模型映射与 API Key 管理 "> <img src="/static/https://cdn5.telesco.pe/file/LC4KT66b27t9a9aaNE5-e8one71QTrNn3e50Xl1crPRkhlgR1VLM9WaUIvh863hCC181dg0FuYlPjWrR0_aysfFzJqKvGE-JSeSJmpc-YaL5U2LV_OBs9eE7N06gDAHBrOmqrxKKwkkvGa1KRXBuVeWNcrulZ7XUHjVzFGjsx-YY6nvk2KpI7WXk16npJnuKnkrUzsyO8tYUKKsqF7Sv4aP2uCQenzT4AeOMUsyCU4iOmr50N_w12Rl9CIveXmfazGE9YaJMDN7aEaAI9VxwmO9X2IszhfzqIx-vsuTJ1OWW-IMcQQJaZ1SjN5wYs1S1ykD8AQMODClce0mDn7fMgA.jpg" alt="Squirrel: 企业级 LLM 网关, 统一代理 OpenAI、Anthropic 及兼容 API• 一次接入统一 API, 兼容 OpenAI Chat、OpenAI Responses 与 Anthropic Messages, 并支持请求/响应/流式互转 • 智能路由与高可用, 支持规则路由、负载均衡、自动重试与故障转移, 可按成本或优先级自动选模型 • 全面可观测性 + 管理面板, 记录 Token 用量/延迟/成本与请求日志, 提供 Next.js 管理后台做供应商、模型映射与 API Key 管理 " width="226" height="357" loading="eager" /> </button> <div class="modal" id="modal-4854-1" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-4854-1" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-4854-1" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><div class="tgme_widget_message_text js-message_text"><u>Squirrel: 企业级 </u><mark class="highlight"><u>LLM</u></mark><u> 网关, 统一代理 OpenAI、Anthropic 及兼容 API</u><br /><br />• 一次接入统一 API, 兼容 OpenAI Chat、OpenAI Responses 与 Anthropic Messages, 并支持请求/响应/流式互转 <br />• 智能路由与高可用, 支持规则路由、负载均衡、自动重试与故障转移, 可按成本或优先级自动选模型 <br />• 全面可观测性 + 管理面板, 记录 Token 用量/延迟/成本与请求日志, 提供 Next.js 管理后台做供应商、模型映射与 API Key 管理 <br /><br /><a href="https://github.com/mylxsw/llm-gateway" target="_blank" rel="noopener" title="https://github.com/mylxsw/llm-gateway">https://github.com/mylxsw/llm-gateway</a><br /><br /><a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23%E7%BD%91%E5%85%B3" title="#网关">#网关</a> <a href="/search/result?q=%23%E4%BB%A3%E7%90%86%E6%9C%8D%E5%8A%A1" title="#代理服务">#代理服务</a> <a href="/search/result?q=%23OpenAI" title="#OpenAI">#OpenAI</a> <a href="/search/result?q=%23Anthropic" title="#Anthropic">#Anthropic</a> <a href="/search/result?q=%23%E5%8F%AF%E8%A7%82%E6%B5%8B%E6%80%A7" title="#可观测性">#可观测性</a> <a href="/search/result?q=%23%E8%B4%9F%E8%BD%BD%E5%9D%87%E8%A1%A1" title="#负载均衡">#负载均衡</a> <a href="/search/result?q=%23%E6%95%85%E9%9A%9C%E8%BD%AC%E7%A7%BB" title="#故障转移">#故障转移</a> <a href="/search/result?q=%23%E6%88%90%E6%9C%AC%E4%BC%98%E5%8C%96" title="#成本优化">#成本优化</a> <a href="/search/result?q=%23Docker" title="#Docker">#Docker</a> <a href="/search/result?q=%23Nextjs" title="#Nextjs">#Nextjs</a> <a href="/search/result?q=%23Python" title="#Python">#Python</a></div>http://telegram.zhetengsha.eu.org/posts/4827http://telegram.zhetengsha.eu.org/posts/4827Sat, 14 Feb 2026 10:01:22 GMT<video src="/static/https://cdn5.telesco.pe/file/b9c27bc724.mp4?token=GegjyJtYTW4SA8mZ7wmBV_vKJ_7ZOIMVQmdynF68m2vBhtz-FU8YPwkaJWYcM27ahLlHPSynhFjJidf8EZTXQMxDPmWk0LDCgRf6GLB8SF-ysKbPDMYRunet9pwlAmN72H7HZbzvyJ70nknbh7-koWP5Qw3Z58Sp1HJo4j0Pr68_X48Fz6sHcKl44qzCPUE2zORIvmzyrvZBThmTVD2Il7X9qEKF1BN2IVNpnCx5prLjZLK-uC6vpqJ0O-gvcNu6zlkividMvAdJmAZtDPGcPmiA8TxCV4EunOY3-asCcaP9t4N51fqUcaRdzWKbZKBSkF91w5cd2Z7HxODXLv_hlQ" class="tgme_widget_message_video js-message_video" width="100%" height="100%" preload="auto" muted autoplay loop playsinline controls webkit-playsinline></video><u>Reader: 面向 </u><mark class="highlight"><u>LLM</u></mark><u> Agent 的开源生产级 Web 抓取与爬取引擎</u><br /><br />• 内置 Cloudflare/Turnstile 等反爬绕过能力，包含 TLS 指纹、DNS over TLS、WebRTC 掩蔽 <br />• 抓取后直接输出干净的 Markdown/HTML，自动抽取正文并清理导航栏、弹窗、Cookie 横幅等噪音 <br />• 同时提供 CLI 与 Node.js API，支持并发批量抓取、站点 BFS 爬取、浏览器池与代理轮换用于稳定上线 <br /><br /><a href="https://github.com/vakra-dev/reader" target="_blank" rel="noopener" title="https://github.com/vakra-dev/reader">https://github.com/vakra-dev/reader</a><br /><br /><a href="/search/result?q=%23%E7%BD%91%E9%A1%B5%E6%8A%93%E5%8F%96" title="#网页抓取">#网页抓取</a> <a href="/search/result?q=%23%E7%BD%91%E7%BB%9C%E7%88%AC%E8%99%AB" title="#网络爬虫">#网络爬虫</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23Agent" title="#Agent">#Agent</a> <a href="/search/result?q=%23Cloudflare" title="#Cloudflare">#Cloudflare</a> <a href="/search/result?q=%23Nodejs" title="#Nodejs">#Nodejs</a> <a href="/search/result?q=%23CLI" title="#CLI">#CLI</a> <a href="/search/result?q=%23Proxy" title="#Proxy">#Proxy</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a> <a href="/search/result?q=%23%E5%BC%80%E6%BA%90%E5%B7%A5%E5%85%B7" title="#开源工具">#开源工具</a> <a href="/search/result?q=%23%E7%94%9F%E4%BA%A7%E7%8E%AF%E5%A2%83" title="#生产环境">#生产环境</a>http://telegram.zhetengsha.eu.org/posts/4767http://telegram.zhetengsha.eu.org/posts/4767Sat, 07 Feb 2026 10:01:02 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-4767-0" popovertargetaction="show" aria-label="Open image preview: Voratiq Agent Leaderboard: 面向真实工程任务的 AI 代理对战排行榜• 基于真实工程规格让多个 agent 正面对比，人工复核结果并合并最优方案 • 用评分与 90% 置信区间量化表现，同时展示任务用时分布与效率 • 提供可视化散点图与完整排名，便于快速定位高质量模型与性价比选项 "> <img src="/static/https://cdn5.telesco.pe/file/PbDV4d2sCRq-RJF8jTfgGPaCMAPVxIonIes12j0oeycyubV0n1t_XSy_wK0IcMVzNyikIu5dRaKbLoCyeHMmqDfzlQHRb2PxO5BqaRFHG7Rbr6_1nDQXfQeZS9nlE3LsRBosDw4Y2zY9gDXkhGKQOFfUTNaHzSBedlY-ODvIQcdGs5HPWIQrtSad3zdS2Kr2ny4bbYTCvnfGmqGGCadpu5svak5k8dz4l8OPWsMjl1l_u1VvttK_rAJ2zGNhAcuR8kqNPPH5FNy1gfoDJFfV2q7CPY3F0JVXpdD5ouWh4vwSBURIroFRTsd9mSdwMtG0mkaEuP_iTOE3sJT8jshIOQ.jpg" alt="Voratiq Agent Leaderboard: 面向真实工程任务的 AI 代理对战排行榜• 基于真实工程规格让多个 agent 正面对比，人工复核结果并合并最优方案 • 用评分与 90% 置信区间量化表现，同时展示任务用时分布与效率 • 提供可视化散点图与完整排名，便于快速定位高质量模型与性价比选项 " width="800" height="636" loading="eager" /> </button> <div class="modal" id="modal-4767-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-4767-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-4767-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>Voratiq Agent Leaderboard: 面向真实工程任务的 AI 代理对战排行榜</u><br /><br />• 基于真实工程规格让多个 agent 正面对比，人工复核结果并合并最优方案 <br />• 用评分与 90% 置信区间量化表现，同时展示任务用时分布与效率 <br />• 提供可视化散点图与完整排名，便于快速定位高质量模型与性价比选项 <br /><br /><a href="https://voratiq.com/leaderboard" target="_blank" rel="noopener" title="https://voratiq.com/leaderboard">https://voratiq.com/leaderboard</a><br /><br /><a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23Agent" title="#Agent">#Agent</a> <a href="/search/result?q=%23%E6%8E%92%E8%A1%8C%E6%A6%9C" title="#排行榜">#排行榜</a> <a href="/search/result?q=%23%E5%B7%A5%E7%A8%8B%E6%95%88%E7%8E%87" title="#工程效率">#工程效率</a> <a href="/search/result?q=%23%E5%9F%BA%E5%87%86%E6%B5%8B%E8%AF%95" title="#基准测试">#基准测试</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a> Copilot <a href="/search/result?q=%23Claude" title="#Claude">#Claude</a> <a href="/search/result?q=%23Gemini" title="#Gemini">#Gemini</a> <a href="/search/result?q=%23OpenAI" title="#OpenAI">#OpenAI</a>http://telegram.zhetengsha.eu.org/posts/4679http://telegram.zhetengsha.eu.org/posts/4679Wed, 28 Jan 2026 01:03:32 GMT<div class="image-list-container image-list-odd"> <button type="button" class="image-preview-button image-preview-wrap" popovertarget="modal-4679-0" popovertargetaction="show" aria-label="Open image preview: LM Speed: 一键测出你的 LLM API 到底快不快当然我知道大多数人是去看有什么 API 可以白嫖的... "> <img src="/static/https://cdn5.telesco.pe/file/ixLjE3f2oCg9Lh91wWi6LcLPlH6C1PPATMkqlAmyvx0hcKdVpHaTba0G7KlbXlpZUycaaO25MWZMGZ9V-twPpsmMYaxFymuiWdm5EfV8oOHYinXBmdHAQqtapJUNPoe7x91p48woYeWoNmzYmK2z1QDyvEpc4D-aoA1R85dMFaUfKHojumlcu5mSelNfAWW3DhuvO_RxT9ZyRGORSy43dtJt9I2Wxb-mzfS3BQGe2AQA3By2HsVxFbQDXV4bdboOnGPUEfkjYVZ-DbYH_KG_UIM9DI3m1l9kvTrzeqtr5oZSLgTy3o0-sFuLQBKQa73Gv7C1skv6mU7xPsBqW3saoA.jpg" alt="LM Speed: 一键测出你的 LLM API 到底快不快当然我知道大多数人是去看有什么 API 可以白嫖的... " width="667" height="800" loading="eager" /> </button> <div class="modal" id="modal-4679-0" popover="auto" aria-label="Image preview"> <button type="button" class="modal__backdrop" popovertarget="modal-4679-0" popovertargetaction="hide" aria-label="Close image preview"></button> <button type="button" class="modal__close" popovertarget="modal-4679-0" popovertargetaction="hide" aria-label="Close image preview">×</button> <div class="modal__surface"> </div> </div> </div><u>LM Speed: 一键测出你的 </u><mark class="highlight"><u>LLM</u></mark><u> API 到底快不快</u><br /><br />当然我知道大多数人是去看有什么 API 可以白嫖的... <a href="https://lmspeed.net/zh/category?q=%E5%85%8D%E8%B4%B9" target="_blank" rel="noopener" title="https://lmspeed.net/zh/category?q=%E5%85%8D%E8%B4%B9">https://lmspeed.net/zh/category?q=%E5%85%8D%E8%B4%B9</a><br /><br />• 多提示词综合测速，量化首字延迟、吞吐（Tokens/s）、总耗时等关键指标<br />• 支持自定义 Base URL + API Key，兼容官方与各类 OpenAI 兼容代理端点<br />• 内置实时排行榜与可视化分析，快速对比不同 Provider / Model 的性能差异<br /><br /><a href="https://lmspeed.net/zh" target="_blank" rel="noopener" title="https://lmspeed.net/zh">https://lmspeed.net/zh</a><br /><br /><a href="/search/result?q=%23%E5%A4%A7%E6%A8%A1%E5%9E%8B" title="#大模型">#大模型</a> <a href="/search/result?q=%23API%E6%B5%8B%E9%80%9F" title="#API测速">#API测速</a> <a href="/search/result?q=%23%E6%80%A7%E8%83%BD%E5%88%86%E6%9E%90" title="#性能分析">#性能分析</a> <a href="/search/result?q=%23%E5%BB%B6%E8%BF%9F%E4%BC%98%E5%8C%96" title="#延迟优化">#延迟优化</a> <a href="/search/result?q=%23%E5%90%9E%E5%90%90%E9%87%8F" title="#吞吐量">#吞吐量</a> <a href="/search/result?q=%23%E6%8E%92%E8%A1%8C%E6%A6%9C" title="#排行榜">#排行榜</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a> <a href="/search/result?q=%23OpenAI" title="#OpenAI">#OpenAI</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a>http://telegram.zhetengsha.eu.org/posts/2265http://telegram.zhetengsha.eu.org/posts/2265Wed, 08 Jan 2025 01:01:54 GMT<u>Darwin Eligibility Override</u><br /><br /><blockquote>项目旨在不禁用系统完整性保护 (SIP) 或者仅禁用一次的情况下 实现永久在任意 Mac 上使用 Xcode <mark class="highlight">LLM</mark> / Apple Intelligence。</blockquote><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6329896824882597500.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://github.com/Kyle-Ye/XcodeLLMEligible" target="_blank" rel="noopener" title="Kyle-Ye/XcodeLLMEligible">Kyle-Ye/XcodeLLMEligible</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327880277607583303.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/2" target="_blank" rel="noopener" title="频道资源合集">频道资源合集</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327792445526380187.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha/957" target="_blank" rel="noopener" title="恰饭推荐">恰饭推荐</a> <img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6021860481923288796.webp" alt="" loading="lazy" width="20" height="20" /> <a href="https://t.me/zhetengsha_group" target="_blank" rel="noopener" title="群组">群组</a><br /><br /><img class="tg-emoji" src="/static/https://telegram.me/i/emoji/6327836885552991740.webp" alt="" loading="lazy" width="20" height="20" /> <a href="/search/result?q=%23mac" title="#mac">#mac</a> <a href="/search/result?q=%23macOS" title="#macOS">#macOS</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23ChatGPT" title="#ChatGPT">#ChatGPT</a> <a href="/search/result?q=%23Intelligence" title="#Intelligence">#Intelligence</a> <a href="/search/result?q=%23LLM" title="#LLM">#LLM</a><a class="tgme_widget_message_link_preview" href="https://github.com/Kyle-Ye/XcodeLLMEligible" target="_blank" rel="noopener" title="Contribute to Kyle-Ye/XcodeLLMEligible development by creating an account on GitHub."> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - Kyle-Ye/XcodeLLMEligible" src="/static/https://cdn4.telesco.pe/file/tiLT6zS6bA2zfHjQxCK3rOVDKuflkfB423T_0ABVMDj32X_IZiPXfNya8wy7zNnfXN3Y6LHHjYVm3QmjHXMEIPgkVvnr5yaJFTF3aRyKoJDj0rLS4v3AHiILUSwkTtbxZcJnnuvSmjWrmNGOmfIdXDwAaenusS2LgXa38SyI3YQUvIUiTIpVqI8xdn-rPcHT65gul5NEqSTHvujTDxXtc31JEegrXu2Fk_NU_CnrFvGsvy7_iKcieM1skXKOGvpyzvH4u5_cq5KlYqSnLUF3DNxCBzk9Uq-sJ6FysvKcl0siWvRViQACTS4AyKmynPWGAbQ1BdtcSV2LRr8Bjgky1w.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - Kyle-Ye/XcodeLLMEligible</div> <div class="link_preview_description">Contribute to Kyle-Ye/XcodeLLMEligible development by creating an account on GitHub.</div> </a>