@checkdotfast · X

dev_dan_2 · HN

WatchMachineGo – A visualizer to show hardware performing LLM inference

@RealAkrazo · X

@honitec · X

做了一个每天自动刷新的模型定价 API 替代 LiteLLM，每天拉最新的推理定价 为什么需要「每天」？ 因为 AI 模型价格一周一变 3月 $8/M → 6 月 $3/M，半年跌了一半多 半年前到处喊 GPU 短缺 现在全员降价抢客户 估值故事讲的是稀缺 现实是价格比成本跌得快 这中间的落差，就是泡沫

@apptheism · X

Shipped Starwattage ⚡️ ( A dynamic pay-to-rank board for indie startups. Projects burn "Wattage" over time to fight for rank #1. Free wattage slots open for early builders to kickstart visibility!

@Sensha_biyori · X

@Amit63390 · X

Tomas_Metcalfe · HN

I've built a fitness app that derives exercise kinematics from ordinary video I ran Reddit ads recently. They generated very good click-through rates, but essentially no signups. I will be honest, I had not given onboarding a single thought till the app was mostly built… Hence I put together the landing page, demo and map app with training spots and “bar bounty” leaderboards, these parts were made with AI tools. The rest was a 2.5 year slog built on decades of coding, sport science , and AI sinc

pepperpoppins · HN

Trunchbull, run real models against any benchmark in your browser

plutozc · V2EX

做了两个 MiniMax H3 AI 视频生成入口 最近在用 MiniMax H3 做一些 AI 视频的快速试验，发现自己常在文字、图片参考、画幅和时长之间来回切，所以顺手做了两个偏不同工作流的入口，想听听大家会不会觉得这种拆分有用。 一个是偏快速迭代的版本： [H3 Max]( https://h3-max.org) 主打从文字或参考图生成短视频，重点放在快速生成 768p 片段和提示词里的画面、镜头、音频描述。 另一个是功能更完整的工作台： [MiniMax H3 Max]( https://minimaxh3max.org) 同样支持 text-to-video 和 image-to-video ，可以按需要选 768p 、2K 或 4K ，也能调时长和横竖画幅。 目前都还是刚上线的个人项目： - 文字或图片作为起点生成视频 - 一个提示词里写画面、动作、镜头和声音方向 - 方便先快速出一个版本，再回头改提示词 两站都不是官方站，生成结果和速度会受模型设置及队列影响。现在页面和体验还在继续补，尤其想听听大家更在意“快点出第一版”，

@AmitGajbhare · X

We build - a modern typing platform that helps people improve typing speed and accuracy with AI-powered practice, real-time feedback, progress tracking

cdnsteve · HN

Tokenstead, find AI models for your hardware