A repository for cody. Contribute to juancgarza/cody development by creating an account on GitHub. Read more ›
From pretraining to RLHF/GRPO — every algorithm hand-written in pure PyTorch. Read more ›
Navigation when there is no GPS and no Internet. Contribute to deepanwadhwa/anumaan development by creating an account on GitHub. Read more ›
Cross-platform launcher for local AI CLI and desktop tools - tjbmoose09/ai-tool-launcher Read more ›
Improve your productivity with AI-powered automatic time tracking. No manual start/stop buttons. Track time automatically with local AI analysis and generate daily work logs. Read more ›
A local-first database of your academic papers built to support arXiV papers and more - linxiv-dev/linXiv Read more ›
Hotkey-triggered screenshot injection into your last active terminal session - kubestellar/hotshot Read more ›
| | | | | - | - | - | | | **Hacker News**new \| past \| comments \| ask \| show \| jobs \| submit | login | AI-gateway product that cuts LLM API TOKEN costs by 40-70% 1 point by arnab777 11 minutes ago \| hide \| past \| favorite \| discuss Guidelines \| FAQ \| Lists \| API \| Security \| Legal \| Apply to YC \| Contact Read more ›
Create personal Mac apps instantly with a prompt. Supports on-device and cloud LLMs - Jeidoban/Ironsmith Read more ›
lightweight, tmux-backed meta-harness for coding agents - mupt-ai/relaymux Read more ›
One API call turns any web page into LLM-ready Markdown. Built for AI agents, RAG pipelines and scrapers. Fetch, render JavaScript, strip the clutter. Free to start, plans from $29/mo. Read more ›
Chromium Embedded Framework (CEF) for SwiftUI on macOS. Three hosting modes (Alloy NSView, Chrome runtime window, OSR/Metal IOSurface), automated CEF download + bundling via SwiftPM plugin. - Rajan... Read more ›
Athena is a desktop control surface for AI coding agents with shared project context. Built with Electron, React, and FastAPI, it embeds native terminals for Codex, OpenCode, Claude, and Hermes wit... Read more ›
I rented an 8×B200 and tried to run GLM-5.2 on TileRT, the runtime MiMo used to push a 1T model past 1000 tok/s. TileRT doesn't support GLM-5.2, so I reverse-engineered its IndexShare attention and weight-remapped it onto the kernel. Result: ~480 tok/s, OpenRouter-identical quality, ~5× faster than vLLM on the same GPUs, capped at 2048 tokens by the closed kernel. The apples-to-apples numbers (TileRT 480 / vLLM 96 / OpenRouter 104), why MTP made vLLM slower, and what it'd take to beat it. Read more ›
Fonos - voice AI assistant. Contribute to ethannortharc/fonos development by creating an account on GitHub. Read more ›
Draft keeps your team's AI sessions grounded. It runs in the background, capturing context from multiple sources, then injects it for everyone. - idodekerobo/draft Read more ›
Quick Access for Pass. Contribute to CiTroNaK/Quick-Access-for-Pass development by creating an account on GitHub. Read more ›
Tiny ruby terminal for X11. Contribute to vidarh/rubyterm development by creating an account on GitHub. Read more ›