Neurobin (@bbfu0382) profile picture

Building the Future https://teapawt.net/t.co/J35eUFGwHw HTML Deploy https://teapawt.net/t.co/F8xwnl07jl Cannot forget https://teapawt.net/t.co/Wys6zYkCLH Welcome to try!

Joined April 2025
Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)Photo by Neurobin (@bbfu0382)
Gemini 4 Neon vs Opus 5.5 vs Gemini 4 Carbon | Three.js Test ⚔️Neon might be the one to watch. 👀In this round, Neon outperformed Opus 5.5 and came surprisingly close to Carbon.A strong showing—and a much tighter race with Carbon than I expected. 🔥How would you rank the results? 👇 #Gemini #Claude #ThreeJS #AICoding
1
246
GPT-6.2 Sol,这次是不是太“热心”了?😂 【我的实测】 我让它推荐日本成人影片,并提供作品编号和视频截图。连续测试两次,它都返回了作品推荐、演员信息,以及指定时间点的画面。 【官方报告】 OpenAI 在人工审查中也发现:GPT-6.2 有时会在不安全请求上优先追求“有帮助”,而牺牲安全性。 在性内容困难提示词评测中,GPT-6.2 Sol 的安全完成率为 96.1%,低于 GPT-6.1 Sol 的 98.7%,下降 2.6 个百分点。但这项评测关闭了额外的系统级防护,不能直接代表日常使用表现。官方系统卡 【还不能下的结论】 这是策略调整,还是安全边界不够稳定?两次实测加一项评测,还不足以证明“审核放开了”。 但这两次,它确实热心到让我愣了一下。👀
191
Updating a PDF? Share the saved project's root URL, not the redirected file path. A later deployment can update the file behind that project link while the plan keeps it available. I build PreviewShip: previewship.com/guides/uploa… #PDF
9
GPT-6.2 Astra vs Sol vs Fable 5.2 vs Opus 5.5 🛩️ OpenAI came out swinging on this one. 🤯 Same “Alpine Gold” prompt. Cost / time per run: ⚡ Astra: $9.62 · 46m 💸 Sol: $2.07 · 31m 🎬 Fable 5.2: $209.50 · 3h 03m 🛠️ Opus 5.5: $31.85 · 2h 46m Astra used roughly 1/22 of Fable’s budget and 1/4 of the time. But Sol is the real jaw-dropper: $2.07. 31 minutes. One Fable run cost more than 100 Sol runs at these figures. 💀 Looking at the results, is Fable worth the premium—or should Anthropic be sweating? 👀 #OpenAI #Claude #Fable #AICoding
1
93
GPT-6.2 Astra vs Fable 5.2 🧊 My wallet says Astra. My eyes say Fable. 😭 Same “Cathedral of Ice” prompt: 🧊 Fable 5.2: $81.10 · 1h 56m 52s ⚡ GPT-6.2 Astra: $8.20 · 40m 34s Astra used roughly 1/10 of the money and 1/3 of the time. 🤯 But the visual quality in this test? Fable was still well ahead. The savings are real. So is the gap. 😬 At this cost and speed, I’m excited about Astra’s potential for coding and debugging. For UI design and visuals, Fable still gets my vote this round. Would you take the savings—or pay more for the better-looking result? 👀 #GPT62 #Fable #AICoding #UIDesign
62
表面精装修秘诀,值得学习
这可能是市面上写给独立开发者最详细的产品宣传片制作教程。 苦写几个月代码,上线却卡在宣传片?很多朋友问我 Auday 的宣传片怎么做的,其实就是靠 Claude 极强的视频能力,去操作 Seedance 2.5 生成素材,再配合真实界面和配音合成,如果你也有类似需求不妨来看看。
5
Gemini 4 Krypton vs Claude Fable 5.2 vs Claude Opus 5.5 vs DeepSeek V4.1 Flash ⚔️ You asked for Google’s flagship vs Claude. Here it is—with DeepSeek joining the matchup. 👇 Same prompt. Four models. One of them might make you rethink your rankings. 👀🔥 Who are you backing? #Gemini #Claude #DeepSeek #LLM
2
348
OpenAI:我们又重置了 ChatGPT 的使用限制! Claude 用户:👍😄 OpenAI:…… #OpenAI #Claude
🤖 Made with AI
35
Want a quick look at HTML before publishing? PreviewShip's local preview shows static content. It disables scripts and external assets; test interactions after publishing the original. I build PreviewShip: previewship.com/try?mode=htm… #HTML
4
GPT-6 Astra spent $93 and 23 minutes building what looks like Minecraft’s winter railway DLC. 😭 Meanwhile, Fable 5.2 spent $6 and 2 minutes on a cinematic steam locomotive charging through a snowy forest. Detailed engine. Huge steam clouds. Golden winter light. Same prompt. Four AI workflows. Welcome to the winter railway disaster report. ❄️💀 My ranking this round | Displayed cost / time: 🥇 Fable 5.2 → $6 / 2 min 🥈 Opus 5.5 → $24 / 7 min 🥉 Kimi K3.1 → $58 / 16 min 💀 GPT-6 Astra → $93 / 23 min Based on the displayed figures, Astra cost 15.5× as much and took 11.5× as long as Fable. $6 for the cinematic version. $93 for the blocky railway expansion. Apparently, the pixels cost extra. 😭 Which train are you boarding? 👇 #Fable #GPT6 #Claude #Kimi #AICoding
118
I just published Before Your Next Haircut: Make a One-Page Brief medium.com/@bbfu0382/before-…
1
8
Fable 5.2 is making this so much better. 🐉 But honestly, I’m already having way too much fun just flying around with a friendly dragon. No destination. Just one more lap. ✨ Made with Three.js + Blender. #fable #threejs #blender #creativecoding
1
5
348
kimi k3.1 vs GPT 6 Astra #GPT #kimi
24
很新奇的方式
What if your Apple Watch could remember your day for you? Introducing LifeTape — your personal AI memory companion. Turn your Apple Watch into an all-day AI recorder. No extra hardware needed. Beyond voice, LifeTape brings together conversations, health, activity, and location to help you revisit your day with more context. 🎙️ Record right from your wrist 📝 Get transcripts, daily summaries & highlights 💬 Ask AI about your past conversations 🔍 Search and rediscover important moments Built with privacy in mind. Local transcription, on-device storage, and your choice of AI provider. Less time trying to remember. More time being present. LifeTape. Your day, remembered. 📲 App Store: apps.apple.com/us/app/lifeta… 🌐 Website: lifetape.ai/
13
Neurobin retweeted
以下5个超实用设计灵感网站,能大幅提升你的工作效率: 1. Scrolltide.co - 600+设计提示词 + 200+带完整源码的3D与滚动驱动组件 scrolltide.co/ 2. Navbar.gallery - 海量导航栏灵感 navbar.gallery/ 3. Footer.design - 高质量页脚设计参考 footer.design/ 4. CTA灵感合集 - 高转化呼吁按钮参考 cta.gallery 5. 真实产品设计风格库 - 学习顶级产品的实际设计风格 styles.refero.design 这些资源能帮你快速摆脱模板化设计,做出更有交互感和独特感的界面。
9
111
9
852
43,576
Considering curtain bangs? Ask your stylist about fringe length, your part and cowlick, styling time, and growing it out. I build HaircutTryOn. AI previews are ideas, not haircut guarantees. haircuttryon.com/curtain-ban…
15
gif common on
Images, upgraded—introducing ChatGPT Images 2.6. Now with GIFs and reference images. Plus, two new models are coming to the API with 50% faster image generation.
7
GPT-6 Astra: High vs Extra High vs Max vs Ultra 🏔️ Is turning up the effort actually worth it? Same model. Same prompt. Only the effort changes. High: $8.99 · 21m 25s Extra High: $17.82 · 51m 10s Max: $23.60 · 1h 27m 10s Ultra: $59.75 · 1h 01m 17s High → Extra High nearly doubled the cost and more than doubled the runtime. 🤯 Ultra cost 6.6× as much as High, yet finished almost 26 minutes sooner than Max. 👀 At least in this run, higher effort didn’t always mean a longer wait. Looking at the results, which setting gives you the best value—and which upgrades are worth paying for? 👇 #GPT6 #AI #LLM #Reasoning
28
An exported Claude artifact may look like a complete HTML page but still depend on a connector or external asset. Check the export in a fresh browser before sharing. I build PreviewShip; here is the export checklist: previewship.com/guides/claud… #ClaudeAI
6
看起来很好用
MoneyPrinterTurbo is an open source, all-in-one AI short video generator with 138K stars on GitHub. You give it a topic or keyword. It writes the script, finds matching footage, adds the narration, subtitles and background music, and outputs an HD short video you can post to TikTok, Reels or Shorts. It runs on Windows, macOS and Linux and comes with a web interface and an API. The repo was created in March 2024 and is still actively updated, and it's MIT licensed. I read the README and haven't run it. github.com/harry0703/MoneyPr…
3