A benchmarking website for open-weight models
Project Details
Live results, coding benchmarks included, agentic benchmarks included, domains and other classifications filters included. (Spoiler: DeepSeek-V4-Vision-Exp rules, but other models have their rule areas): https://beta.locallm.top Evaluated by domain experts (my friends mostly; coding part is evaluated solely by me), classified by domains/languages/intents (can be filtered on the home page), agentic coding benchmark included. This is only public part of the data (whoever has private queries and ev
Similar Projects
Building an app has become easy. Deploying it properly is still painful. So I built something for th
The Deployer puts AI-built apps into your own AWS, Google Cloud, Azure or DigitalOcean acc...
Looking for alpha testers for setu-ts: An open-source, plugin-aware TypeScript development framework
Plugin-first enterprise backend framework built on Hono.
I built a free tool for catching all internet startup trends and domain suggestions which will be us
BuyHype watches emerging companies, technologies, products, communities and conversations...
BetaReward - A marketplace for Google Play closed testing
BetaReward is a global app beta testing platform: publishers launch Closed Testing tasks w...
Just launched an audio tour guide app for road trips.
Whistlestop is a GPS audio tour guide app for iPhone. It narrates the landmarks, nature, h...
Cloud-based Coding Plugin: Real-time Optimization Suggestions from AI Models for Your Web Pages
工作台开任务、跑云端 Agent，浏览器插件对着真实页面提优化建议。租户 RBAC、GitLab SSO、KYC...
