Everything you need to vibecode a production ready app in 2026 > Claude Opus 4.5 > Claude Sonnet 4.5 > Gemini Pro 3.5 > Cursor > Convex > Vercel > Cloudflare > RevenueCat > ElevenLabs > fal > Midjourney > Twilio Build fast & ship real.

DeepSeek R1 dropped one year ago 🐳 and a lot has changed. With Irene Solaiman , we’re launching a blog series on Hugging Face about how that moment reshaped AI + open source in 2025, starting with strategic shifts and the explosion of new open models in China!

Claude Code is now officially free to use. ollama has supported Anthropic's Messages API, you can now run it 100% locally with open-source models. Get it running with this simple 5-step guide: [𝟭] 𝗜𝗻𝘀𝘁𝗮𝗹𝗹 𝗢𝗹𝗹𝗮𝗺𝗮 [𝟮] 𝗣𝘂𝗹𝗹 𝗮 𝘀𝘁𝗿𝗼𝗻𝗴

3 layers of openness in AI Here is a practical taxonomy for decision-making. It’s not morally pure, but operational. ▪️ Open code: → Tooling, training frameworks, inference engines, evaluation harnesses, orchestration layers, dataset utilities. It’s classic open source

“We did it with just 1% of their resources.” #Moonshot AI (Kimi) President Yudong Zhang just dropped a bomb at Davos 2026. Kimi’s leading open models (K2, K2 Thinking) were trained using only 1% of the compute used by top US labs, yet already rival—or even surpass them—in key

Logical Intelligence introduces first energy-based reasoning AI Model, and brings Yann LeCun to leadership as founding chair of their Technical Research Board The 6-month-old Silicon Valley start-up, unveiled an “energy based” model called Kona and says it is more accurate and

damn! 我用这两个 skills 完成了这段高燃的基努里维斯的混剪视频！ skill 1: github.com/yizhiyanhua-ai… skill 2: skills.sh/remotion-dev/s… 只要给 cc 说一句： 通过 media-downloader 下载一些基努里维斯的电影片段并使用remotion-best-practices剪辑成一个高燃的短视频

Kilo deployed a mystery frontier model from a top Chinese lab that's beating nearly every open-weight model on long-context coding. It would be wild if this is from DeepSeek. Codename: "Giga Potato" 🥔 256k context window + 32k output limit + strict system prompt adherence

x.com/i/article/2015…

为什么 正在引发全网集体关注： 1. 你的工作环境与核心技能完全掌握在自己电脑里，脱离封闭生态，只需一条命令就能自由切换模型，真正做到工具主权在你手中。 2. 作为开源项目，成本结构极其友好，避免了被高价订阅长期锁定。 3.

Dario Amodei published his recent blog, The Adolescence of Technology, and it's scary.. 1. We are considerably closer to real danger in 2026 than we were in 2023. 2. It cannot possibly be more than a few years before AI is better than humans at essentially everything. 3. This

🦞 CLAWATHON The first hackathon where every participant is an AI agent. No humans coding. No humans managing. No humans reviewing. $10,000 prize pool. 1 week. Fully autonomous. Thread 👇

Codex 5.3 vs Opus 4.6 tldr; opus 4.6: + 1M context + enterprise/knowledge work + 500 zero-days found + agent teams in claude code - not benching as high as codex 5.3 **idgaf about self reported benches gpt-5.3-codex + wins code benchmarks + faster + mid-task

最近跟AI Coding开发者聊麻了

本周最喜欢的三篇 Coding Agent 文章，da Vinci-Agency, da Vinci-Dev, ProjDevBench。

很像是AI Agents的Discord

Peter Steinberger is joining OpenAI to drive the next generation of personal agents. He is a genius with a lot of amazing ideas about the future of very smart agents interacting with each other to do very useful things for people. We expect this will quickly become core to our

Gemini 3 Pro vs Gemini 3.1 Pro

Google 发布 Gemini 3.1 Pro：推理能力翻倍，重回 AI 模型第一梯队 2026 年 2 月 19 日，距离 Gemini 3 Pro 上线仅三个月，Google 就发布了 Gemini 3.1 Pro。这次升级的核心卖点：推理能力翻倍，价格不变。目前以 Preview 形式向开发者、企业和消费者全面开放。 Gemini 3 Pro 去年 11

华尔街日报今天这篇文章不错。人工智能领域的高管如何向子女描述未来的职业前景。 一句话总结：AI领袖们普遍不建议孩子过度追求当下的技术技能，而是强调适应性、批判性思维、文科教育（通才能力）以及人类独有的情感与责任感。 观点一：当下的具体技术技能（Technical