milk
477 posts
月之暗面旗下的 kimi-cli 正在进行架构重写并更名为 kimi-code。
为解决原 Python 版本冷启动慢、分发体积大等瓶颈，团队全面转向 Anthropic 旗下 Claude Code 的技术范式，基于 TypeScript 与 Bun 运行时完成重写，实现了毫秒级冷启动与高度流畅的 React Ink 终端界面。
重构的直接诱因是原 Python 版本在终端分发与响应体验上的局限。开发人员在提交超 3.8 万行代码的合并申请（PR #1707）时，幽默地在 PR 描述中写道「kimicli 用 Python 是彻底的失败，立刻重构为 TS」，宣告 Kimi 彻底告别 Python 技术栈。
新版支持将屏幕录像拖入终端进行多模态分析，并引入计划模式（Plan Mode）光标交互编辑、Emacs 快捷键、双击 Ctrl + C 安全退出等交互细节。
为了保护已有 8700 多个 Star 的开源资产，月之暗面选择维持原 kimi-cli 仓库活跃。官方通过重定向安装脚本，确保用户运行全局命令 kimi 或更新原客户端时，能自动下载并切换到由 Bun 编译的 TypeScript 单文件程序，实现就地无感升级。
kimi-code 同时开放了第三方大模型 API 自定义接入，支持将工具用作跨模型的统一终端编程网关，不再局限于 Kimi 家族。
Yufan Sheng@amehochan
翻译一下，Kimi 自己基于 Python 写的 kimi-cli，在今天换成了基于 Typescript 和 pi-tui 写的新 kimi-code。 已经在 PUA 对应的研发小哥哥加一些我在 Claude Code 上用得很爽的功能。🤡🤡🤡
中文
@JustinGorya @OfficialLoganK Oh my God, stop praising that dead product called AI Studio. They need a good model - don't praise any of their products, everything they make sucks
English
Logan, I’m sorry to say this, but this is the worst release from Google DeepMind yet. I was really looking forward to it because Gemini 3 was a top-tier model. However, Gemini 3.1 started to feel inconsistent and worse, and I don't have any use case for Gemini 3.5 Flash. It’s very disappointing, but hopefully, things will improve. On a positive note, AI Studio received a nice upgrade, Google Flow is looking good, and Gemini Omni is a very impressive model.
English
Gemini 3.5 feels like the start of a new era for Gemini, we spent the last 2.5 years putting the infrastructure, products, team, etc in place (learning lots of lessons along the way).
The model is the product, please keep the feedback coming!
English
@OfficialLoganK The price and performance of this model are terrible, Google needs to wake up!
English
@Mmilk100 有一说一，虽然我不咋用 Gemini CLI 的，但是作为一个 CLI 它其实还不错的，功能丰富 & 社区活跃。
主要的劣势在于 Gemini 模型不行
中文
很难评 😅这次 Google 开大会，统一了 Antigravity 的各类产品线，然后要让用户从 Gemini CLI 迁移到 Antigravity CLI。
算不算对 Gemini CLI 用户和开源社区的背刺啊
developers.googleblog.com/an-important-u…
中文
Antigravity 2.0 和 Codex 长一个样😂
Google 舍得拉下脸抄一下 UI/UX 也不是坏事，这样不会像 Gemini UI 那样难用了
Google Antigravity@antigravity
Introducing Antigravity 2.0, a new standalone desktop application that delivers fully on that original glimpse of a truly agent-optimized experience. Rebuilt from the ground up with multi-agent teams, scheduled tasks, native voice and one-click integration with other Google products. Learn how to get started with Antigravity 2.0 👇
中文
Gemini CLI 无了， 变成了 Antigravity CLI， 然后为了高性能用 Go 重写了
Gemini CLI@geminicli
Transitioning Gemini CLI users to Antigravity CLI We are unifying our efforts around a single harness and platform, Google Antigravity with four distinct surfaces: • Antigravity 2.0 • Antigravity CLI • Antigravity SDK • Antigravity IDE This will allow us to move faster and give you a streamlined experience wherever you do your best work. Rebuilt in Go for speed, Antigravity CLI is available today and brings robust multi-agent orchestration and asynchronous workflows to your terminal. Important things to know: 1. If you are using Gemini CLI through your Google one account (Google AI Pro or AI Ultra) or through Gemini Code Assist for individuals (free offering) we will be helping you migrate your workflows over the next 30 days. 2. No action required for Enterprise users. Enterprise plans and API keys will continue to be supported in Gemini CLI. Read the full details in our blog post → goo.gle/4eWkUgK
中文
Cursor 手握那么多的编程数据，搞这种模型就算达不到 SOTA，感觉也弱不到哪里去。
Cursor@cursor_ai
Introducing Composer 2.5, our most powerful model yet. It's more intelligent, better at sustained work on long-running tasks, and more reliable at following complex instructions. For the next week, we’re doubling the included usage of the model.
中文
Gemini 3.1 Pro - Frontend
Gemini 3.2 Pro - Backend
leo 🐾@synthwavedd
For everyone asking about GPT-5.6's frontend abilities - it still sucks. Nothing is certain in life, except death, taxes, and GPT models generating the sloppiest UIs of all time. Gemini 3.2 Pro is heading in the same direction too, regressing versus 3.1 Pro.
Indonesia
@laozhang2579 gemini最大问题不是智商不够。而是幻觉严重，注意力也差，上下文长了就白瞎。感觉这是架构问题。他的智商永远只有第一发好使，后面就得开新会话。只适合不需要多轮上下文的场景。
中文
@Gc_qube @HarshithLucky3 They really enjoy polishing AI Studio, that almost-unused garbage product, and they're quite smug about it.
English
@HarshithLucky3 ai studio build
stitch
gemini cli
antigravity
> and all of this is useless shit. they need to stop chasing quantity and start focusing on quality. gemini is lagging behind
English
