New in Claude Design: it stays on brand with your design system across projects, lets you edit directly on the canvas, s...
关联讨论 1 条Claude:Blog(网页)Frontier doesn't only mean Anthropic and OpenAI anymore. I built this black hole simulator by simply dropping an illustr...
It's now easier to move local agents to the cloud so they can keep working with your laptop closed. Prompt Cursor from y...
We built an internal AI system called Builderbot. It coordinates agents across our entire codebase. Engineers tag it in ...
.@grok Build is on DigitalOcean Marketplace. 🆕☁️🤖 SSH in. Run grok. Build. @xAI's coding agent, pre-configured on Digi...
BREAKING: GLM-5.2 is now 1st on Design Arena. With an Elo of 1360, GLM-5.2 has jumped ahead of the now unavailable Claud...
Matt Pocock(Total TypeScript 作者)开源了 skills v1,将技能描述的 Token 成本降低 63%。该工具包将技能分为模型可调用和用户可调用,新增 /codebase-design、/domain-modeling、/grilling 三项技能;重写 /writing-great-skills;将 /diagnose 更新为 /diagnosing-bugs 并改为模型可调用;新增 /ask-matt 路由技能,帮助 AI 自动判断时机触发合适工程流程。主推文评价其将 prompt 从咒语拆解为纪律性流程。
Announcing mattpocock/skills v1 - Achieved a 63% reduction in token cost for skill descriptions - Split skills into mode...
1. as a mental model it is more correct to think of fable+ class models as english -> code interpreters - converts your ...
Love this direction from OpenAI. That's what we call being open. In practice, though, many third-party APIs are still Ch...
OpenAI 格局大了 宣布Codex (包含 App 客户端、命令行 CLI 和开发包 SDK) 支持直接接入任何开源大模型 不强制绑定 OpenAI 自家的模型 并且放出了一个文档:手把手教开发者如何把 Codex 客户端底层的“大脑”,替换成免费的开源模型…
Reminder that you can use the Codex App, CLI and SDK with any open source model, not just with OpenAI models. https://de...
Code like a real G😎 Congrats to @Zai_org 's GLM 5.2 ranks #1 as available model on CodeArena 💪 SiliconFlow is proud to...
关联讨论 9 条公众号:智谱(GLM)X:Testing Catalog (@testingcatalog)Hacker News 热门(buzzing.cc 中文翻译)Simon Willison 博客The Decoder:AI News(RSS)X:智谱 Z.ai (@Zai_org)Nathan Lambert:Interconnects(RSS)智谱:研究(网页内嵌数据)Hugging Face:Blog(RSS)GLM-5.2 正式发布,实测显示其 Agent 能力有质的变化。该模型能将地图数据内化到 1M 上下文中,直接知道换电站位置,全程未调用搜索函数,在测试的 20 多个模型中唯一能做到。后端 Agentic Coding 能力提升至总榜第二名。短板是空间理解:虽记住换电站位置,但无法根据当前位置推理最近站点。
Claude Code之父分享团队核心理念:CLAUDE.md越短越好,他自己的仅两行(提PR自动合并、发审批频道),并建议定期清空重写。模型能力每代都在飞涨,人为堆配置、做UI很快会过时,因此坚持CLI而非GUI。终端输出在详细与简洁间通过用户反馈迭代(支持verbose模式)。用AI修bug已很高效,能自行分析heap dump。核心观点是做减法、保持轻,并不断推翻过时使用习惯。
智谱今日正式开源 GLM 5.2,这是首个编程 coding 能力达到 Opus 水平的开源模型。目前该模型已接入 Cola 作为 beta 模型开放测试,定价与官方一致,欢迎体验和反馈。
智谱 GLM 5.2 在编码评测 CodeArena 的可用模型中排名第一。硅基流动同步首发,定价 Input Cache/Input/Output 分别为 $0.26/1.40/4.40 每百万 token,支持 1M 上下文,具备可靠的长时间任务执行能力,编码性能与 Opus 4.8 持平。提供双思考模式:max 侧重深度,high 侧重质量成本平衡。模型完全开源。
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong...
智谱发布并开源 GLM-5.2,定位长周期任务,支持 100 万 token 稳定上下文。引入思考力度控制:GLM-5.2 max 追求极限性能,GLM-5.2 high 兼顾效率。架构采用 IndexShare 机制,每四层稀疏注意力共享 indexer,百万 token 下每 token 计算量降低约 2.9 倍。编码与智能体任务表现显著提升。模型权重以 MIT 许可证开源,API 定价与 GLM-5.1 一致。
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong...
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong...
Cursor AI CEO Michael Truell在X上宣布与SpaceX联手,共同构建有用AI。据推文称,双方合作规模接近600亿美元。年仅二十多岁的Truell被评价拥有多项稀缺品质,主推文认为其未来成就可能并肩甚至超越OpenAI CEO Sam Altman和Anthropic CEO Dario Amodei。目前具体合作细节尚未披露。
Lots to do together. Excited to be joining forces with @SpaceX to build useful AI.
Cursor CEO Michael Truell提出目标:发明一种新型编程,让人用最简洁的方式描述意图,其余交给模型。他反对两种主流设想——“一切照旧”派(继续用TypeScript/Go等正式语言)和“纯聊天机器人”派(对bot说句话就改代码),认为前者低估模型进步,后者缺乏精确控制。他的设想是“代码之后”的世界:用类似英文伪代码表达软件逻辑,能在高层级编辑、精准指向修改,让难懂的符号变成可读可改的形式。
Cursor 发布新产品 Origin,定位为 AI 原生时代的 GitHub。Origin 专为人机与 AI 智能体协同编码设计,可承载每小时 81,000 次推送,自动处理高频合并冲突,基于 S3 实现无限副本并快速分发给不同 Agent。它原生兼容 Git,现有工作流无需改动,形成从 IDE 编辑器、代码托管到智能评审的全链路闭环。Origin 将于 2025 年秋季上线,现已开放候补名单。
We're launching code storage and git hosting. Origin gives teams and agents a place to host, review, and collaborate on ...
刚被 SpaceX 收购的 Cursor 宣布推出代码托管与 Git 协作平台 Origin,目标是将自身从 AI Editor 升级为端到端开发平台,打通写、存、审、合全链路。Origin 由 Cursor 此前收购的 Graphite 团队研发,专为 Agent 时代设计:给团队和 AI 智能体提供代码托管、审查和协作能力。目前官网已开放 waitlist,预计今年秋季正式上线。
We're launching code storage and git hosting. Origin gives teams and agents a place to host, review, and collaborate on ...
GLM-5.2 以 MIT 协议开源权重,支持 1M 上下文窗口。相比 GLM-5.1,在 Coding、Tool use、Reasoning 上明显提升,尤其在长程 Agent 任务(大规模代码实现、自动化研究、性能优化、复杂调试)中更稳定。提供 Max 和 High 两种推理模式,分别侧重极致性能与 Token 效率平衡。API 价格与上一代相同。社区已在 DeepSWE 等基准上验证其能力。此外,Slide 生成、长文档处理和角色扮演等任务也有进步。
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong...
BREAKING: GLM-5.2 is now 1st on Design Arena. With an Elo of 1360, GLM-5.2 has jumped ahead of the now unavailable Claud...
SpaceX has exercised the option to acquire @cursor_ai in an all-stock transaction with the goal of building the world's ...
关联讨论 11 条TechCrunch:AI(RSS)X:Testing Catalog (@testingcatalog)X:阿易 AI Notes (@AYi_AInotes)X:Michael Truell (@mntruell)The Decoder:AI News(RSS)X:宝玉 (@dotey)IT之家(RSS)X:歸藏 (@op7418)Hacker News 热门(buzzing.cc 中文翻译)The Verge:AI(RSS)Ars Technica:AI(RSS)BREAKING: GLM-5.2 is now 1st on Design Arena. With an Elo of 1360, GLM-5.2 has jumped ahead of the now unavailable Claud...
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong...
Cursor/Graphite's @TomasReimers just announced Origin @cursor_ai's long awaited Git competitor, scalable for agent workl...
BREAKING: GLM-5.2 is now 1st on Design Arena. With an Elo of 1360, GLM-5.2 has jumped ahead of the now unavailable Claud...