Published entries across all sections carrying the “Open source” tag, newest first by publication date on this site.
37 entries
Shanghai Innovation Institute released the Nex-N2.5 family on September 9, 2026 — 35B, 397B and 1.6T models, all under Apache-2.0. On the official table the 1.6T Max tops BrowseComp at 92.6, ahead of Claude Opus 5 and GPT-5.6 Sol, and outscores GLM-5.3 and DeepSeek-V4-Pro on automation and tool-use benchmarks.
AutoClip is an open-source video clipper for long-form material: feed it an interview, podcast, lecture or gaming VOD and it mines the subtitles for an outline, scores segments for highlight-worthiness and cuts clips you approve before anything renders. Unlike cloud clipping services, cutting and rendering stay on your machine with your own model keys — fully offline via Ollama. It ships as a desktop app, a Docker stack and a CLI/MCP pipeline, with export presets for Douyin, Shorts and Bilibili.
html-explainer is an open-source Agent Skill for Claude Code, Codex and other coding agents: give it a topic and it runs research, narration, voiceover, subtitles, frame-by-frame rendering and covers end to end, producing a hard-subtitled MP4. Scenes are written in HTML/CSS/GSAP and rendered with a deterministic seek renderer, so audio and animation stay locked in sync. The default edge-tts voiceover is free and everything runs locally.
A first pick and a backup for every layer — framework, starter, database, auth, payments, hosting, testing, monitoring — all with official links and free tiers. Follow it and your fixed monthly cost stays under $20; most months it is $0.
OpenMausBot is an open-source desktop chat app that turns the Claude Code, Codex and Grok CLIs already on your machine into bots in a sidebar: each bot gets its own personality, model, memory and a computer to work on, and asks before it acts. Instead of juggling terminals and vendor sites, you get one entry point over logins you already have, with transcripts kept local.
An Agent Skill from resciencelab's opc-skills collection (Apache-2.0, 1,827 stars) that wraps Product Hunt's official GraphQL API into 11 read-only scripts. Installed in Claude Code, Codex or Cursor, the agent pulls posts, comments, topics and collections without scraping the web — built for competitor research and idea validation.
Inferact, founded by the original vLLM team, wrote a megakernel inference kernel for Google TPUs. Paired with DeepSeek’s DSpark speculative decoding, 16 TPU v7 chips served Kimi K3 at 709 tokens per second versus 452 on GB200 under the same setup, QbitAI reports. The code is open source.
The front-end design skill from Anthropic's official skills repository. It has the agent ground the design in the product and its industry, commit to a small token system, then build and self-review from screenshots, so the result stops looking like every other AI-generated page.
An MIT-licensed open skill that puts hard limits on how an agent draws IP mascots: 4–7 large shapes per silhouette, three colors at most, legible at 32×32. It pitches three directions first, then generates six candidates in one batch. For developers tired of AI mascots that turn to mush at icon size.
An open-source Agent Skill by Hailey for Codex, Claude Code and Cursor. It makes the agent assemble a Three.js building part by part — foundations, timber frame, brick courses, dougong brackets, tiles — instead of fading in a texture. Built for portfolio pieces, tourism sites and event pages.
The usual way to narrate a video is pasting a script into a cloud TTS: per-character pricing, generic voices, and your unpublished text on someone else's server. yovoice runs 7 local TTS models through audio.cpp on your own hardware, clones a voice from a 1-60 second clip, and ships a full record-trim-export workflow plus a CLI and an Agent Skill. Apache-2.0, for macOS and Windows.
Instatic is an open-source, self-hosted CMS from the CoreBunch team that packs a canvas editor, content management, forms, plugins and a publisher into one Bun server. Published pages are plain HTML and CSS with no framework runtime, a good fit for indie developers and small teams who want off Webflow's monthly bill but find WordPress too heavy.
An open-source clean-room reimplementation of the Claude Code agentic framework in Python and Rust for autonomous terminal engineering.
Free online chat access and WebGPU-accelerated in-browser local inference for the open-weights DeepSeek R1 reasoning model.
An open-source AI-native knowledge base and research agent accelerating complex multi-modal document synthesis for enterprise teams.
A framework-agnostic, lightweight open-source AI chat UI library with native WebMCP client tool discovery and execution.
An open-source TypeScript SDK for executing autonomous coding agents inside secure, isolated execution sandboxes.
An open-source sample app from the Firecrawl team that scrapes a webpage and hands it to an LLM to rebuild as a runnable React app, with conversational iteration, a choice of four models, and cloud sandbox previews—handy for replicating a frontend or spinning up a prototype fast.
A self-hostable, TypeScript web scraping toolkit that covers everything from static pages to JS-rendered pages across three engines, with bulk search-result collection, full-site crawling, and LLM structured extraction, all outputting LLM-friendly data.
MonkeyOCR is an open-source document-parsing tool built on a lightweight multimodal LLM that parses PDFs in three stages—layout, recognition, relation—to reconstruct formulas, tables, and reading order into Markdown, with local GPU inference.
Microsoft's open-source AI-oriented quant investment platform strings data processing, model training, backtesting, and portfolio optimization into a single pipeline, with a high-performance data engine and 20-plus mainstream model benchmarks, running the full research-to-evaluation flow with one command.
An open-source translation library that parses and reflows English PDF papers into a bilingual side-by-side document, using an LLM for translation while preserving formulas, tables, and other original layout as closely as possible; usable as a CLI or embedded in other programs.
croc is an open-source command-line file transfer tool that lets two computers on different networks or operating systems send files directly to each other with a single passphrase, end-to-end encrypted, with resumable transfers and support for multiple files at once.
social-auto-upload is an open-source Python tool that uses browser automation to publish videos to eleven platforms—Douyin, Bilibili, Xiaohongshu, and more—in one click, with image-and-text and scheduled posting on some platforms and multi-account management, suited to running a content-creator matrix.
An open-source customer-service bot for Xianyu (Idle Fish) sellers: it uses an LLM for intent recognition to route buyer messages to specialist agents—negotiation, technical, customer service—and answers automatically using conversation context, with a built-in tiered discount strategy for haggling, keeping the store staffed around the clock.
Syncthing is an open-source continuous file-synchronization tool written in Go that transfers files directly, peer-to-peer, between your own devices without going through any third-party cloud service; it runs automatically, works cross-platform, and suits privacy-conscious individual users.
The open-source tool Story-Flicks takes nothing but a story theme and automatically produces a complete short video with AI illustrations, narration, and subtitles; the text model can connect to DeepSeek, Alibaba Cloud, OpenAI, and other providers, with one-command Docker deployment and multi-language output.
Microsoft's open-source course on building AI agents has grown from 10 to 18 lessons, covering concepts, frameworks, and design patterns through deployment and security, each with a lesson writeup and runnable code, translated into 50-plus languages including Chinese.
An open-source audio/video note-taking tool that runs speech recognition locally with FunASR and organizes the transcript into structured Markdown using a local Ollama model, keeping data on your machine — well suited to meetings, interviews, and lecture recordings.
A command-line tool that uses LLMs to turn epub, txt, srt, and other files into bilingual side-by-side editions, supporting multiple model providers with a session-context mode that keeps translated names consistent across the whole book, plus a free machine-translation path that needs no API key.
Pearcleaner is a free, open-source Mac uninstaller that batch-removes apps you multi-select and automatically cleans up leftovers scattered across the disk, with orphaned-file search and trash monitoring—an open-source alternative to AppCleaner.
An open-source music client built with Flutter that covers desktop and mobile; music sources, playlists, and metadata are all plugged in, playback runs locally with no telemetry, and tracks can be downloaded with tags for offline listening.
An open-source file-cleanup tool written in Rust that scans multi-threaded with caching support, finding duplicate files, similar images and videos, empty folders, oversized files, and corrupted files — fully offline, with GUI, CLI, and Android front ends.
Web-Check is an open-source website intelligence tool that shows dozens of data points—IP, DNS, SSL certificate, security headers, tech stack, performance, and more—just from a URL, with zero-config self-hosting suited to site owners' self-audits and security recon.
Lux is a command-line video downloader written in Go that resolves real stream URLs across 45 sites including Bilibili, Douyin, and YouTube, with quality selection, subtitle downloads, batch playlist fetching, and resumable downloads.
Escrcpy is an open-source desktop app built on the scrcpy core that provides Android screen mirroring and mouse/keyboard control through a graphical interface, with wireless connections, multi-device control, and automation workflows, no command-line flags required.
An open-source tool that converts ebooks into audiobooks entirely on your own machine, with multiple built-in TTS engines, voice cloning, and support for 1,158 languages, producing an m4b with chapters and metadata — no cloud service needed, and it runs on as little as 2GB of RAM.