GitHub 2025-11-04 Daily Trending Projects

BettaFish is a modular multi-agent public-opinion analysis engine for enterprises and researchers that aggregates multi-source multimodal data and produces actionable reports via agent-forum collaboration; however, license, dependency and operational details must be clarified to reduce production risks.

Nano-vLLM provides a compact, readable vLLM inference implementation focused on offline, high-throughput use; it mirrors vLLM's API and integrates practical optimizations (prefix caching, tensor parallelism, Torch compile, CUDA graphs), making it suitable for engineers and researchers pursuing local deployment or extensible inference experimentation.

DeepCode is a multi-agent automated coding platform that integrates Paper2Code, Text2Web and Text2Backend workflows to accelerate prototyping, code synthesis and deployment.

Glow renders Markdown stylishly in the terminal, offering both TUI and CLI modes, multi-platform installation options and customizable styles — ideal for quick README and documentation browsing.

Opencode is a terminal-first AI coding agent with a client/server architecture, provider-agnostic LLM support and a strong TUI focus, enabling controlled local or remote deployments for teams and developers integrating LLM workflows into their development process.

Chef is a Convex‑based backend‑aware AI app builder that quickly generates full‑stack prototypes with a built‑in database, realtime UI and background workflows—suited for developers validating ideas and building internal tools.

PyTorch delivers an integrated tensor and autograd platform for research and engineering, combining flexible dynamic graph programming with high-performance GPU acceleration—suitable from prototyping to production deep learning workflows.

AgenticSeek is a fully local, privacy-first autonomous agent that browses the web, writes and runs code—suited for users running high-performance local LLMs.

LocalAI delivers an OpenAI-compatible local inference platform supporting multiple backends and multimodal models, enabling developers and organizations to run offline, controllable, privacy-focused generative AI on consumer-grade hardware or on-premises environments.

MaxKB delivers a one-stop open-source enterprise agent platform combining RAG and workflow orchestration for self-hosted knowledge retrieval and business automation, ideal for rapid deployment of internal QA and customer service.

deepdarkCTI focuses on collecting and organizing available deep/dark-web threat intelligence sources and search methods, offering teams a browsable source catalog and analysis guidance to support intelligence collection and research decisions.

PageIndex converts long documents into semantic trees and uses LLM-driven tree search for reasoning-based retrieval, delivering vector-free, high-precision, and explainable retrieval for professional domains like finance and law.

For creators and researchers, Deep-Live-Cam enables real-time face-swapping and one-click video deepfakes from a single image with GPU and Apple Silicon acceleration; useful for creative demos and live shows but requires strict legal and ethical compliance.

OpenTUI is an experimental TypeScript TUI framework offering core primitives and React/Solid reconcilers, suited for developing and validating terminal components and tooling.

Everywhere is a context-aware desktop AI assistant that senses on-screen content and responds via hotkeys, integrating multiple LLMs and system tools to provide seamless in-workflow intelligent assistance for individuals and teams.