LLM & Inference

AI API 中转站评测排行与切换工具：实测可用率、延迟与价格 · Open-source AI API relay benchmark, rankings & switcher for Claude Code / Codex

Long-term memory runtime for AI agents — plain Markdown as the source of truth, local ranked retrieval, and an independent sleep-time Manage layer. Claude Code and Codex share one store. No API key.

[ICLR 2023] ReAct: Synergizing Reasoning and Acting in Language Models

Generative AI reference workflows optimized for accelerated infrastructure and microservice architecture.

Generalist and Lightweight Model for Named Entity Recognition (Extract any entity types from texts)

Rapid-MLX is an open-source (Apache 2.0) OpenAI- and Anthropic-compatible LLM inference server for Apple Silicon, built on MLX, focused on reliable tool calling for coding agents. Up to 4× faster than Apple's MLX (mlx-lm) on the same weights.

A series of large language models developed by Baichuan Intelligent Technology

Every front-end GUI client for ChatGPT, Claude, and other LLMs

Multi-LoRA inference server that scales to 1000s of fine-tuned LLMs

TextGrad: Automatic ''Differentiation'' via Text -- using large language models to backpropagate textual gradients. Published in Nature.

A collection of scientific methods, processes, algorithms, and systems to build stories & models.

Eagle: Frontier Vision-Language Models with Data-Centric Strategies

High-performance Inference and Deployment Toolkit for LLMs and VLMs based on PaddlePaddle

OpenDeepWiki is the open-source version of the DeepWiki project, aiming to provide a powerful knowledge management and collaboration platform. The project is mainly developed using C# and TypeScript, supporting modular design, and is easy to expand and customize.

De-AI writing skill for any Agent Skills-compatible agent (77+ via the Skills CLI), with native plugins for Claude Code, Codex, Grok Build, and Antigravity. Narrative-architecture repair for fiction, venue-matched rules for professional prose. Based on StoryScope (arXiv:2604.03136).

Code and models for ICML 2024 paper, NExT-GPT: Any-to-Any Multimodal Large Language Model

Your Cheat Sheet for AI Engineering Interview – Questions and Answers.

⚡FlashRAG: A Python Toolkit for Efficient RAG Research (WWW2025 Resource)

🔥🔥High-Performance Face Recognition Library on PaddlePaddle & PyTorch🔥🔥

[KDD'2026] "VideoRAG: Chat with Your Videos"

OML 1.0 via Fingerprinting: Open, Monetizable, and Loyal AI

An unnecessarily tiny implementation of GPT-2 in NumPy.

The official repo of MiniMax-Text-01 and MiniMax-VL-01, large-language-model & vision-language-model based on Linear Attention

🦞 Just talk to your agent — it learns and EVOLVES 🧬.

[CVPR2024 Highlight][VideoChatGPT] ChatGPT with video understanding! And many more supported LMs such as miniGPT4, StableLM, and MOSS.

Andee lives on your phone or Mac and uses any app like you do—no integrations, no root, any model; memory and passwords never leave your device | Andee 住进你的手机或 Mac，像你一样操作任何 App——无需集成、无需 Root、支持任意模型，记忆和密码绝不离开设备。

Official repo for "Mini-Gemini: Mining the Potential of Multi-modality Vision Language Models"

OneTrainer is a one-stop solution for all your Diffusion training needs.

nanoMuse: an open-source personal agent for every device you own. One agent with a name and a look of its own that does things, keeps working while the app is closed, remembers you, and asks before anything you cannot undo. On Android, iPhone and iPad, Windows / macOS / Linux and in the browser; your phone's screen and your computer as its hands.

MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model.

LLaMA-Omni is a low-latency and high-quality end-to-end speech interaction model built upon Llama-3.1-8B-Instruct, aiming to achieve speech capabilities at the GPT-4o level.

Must-read Papers on LLM Agents.

[EMNLP 2023 Demo] Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding

Learn to build your Second Brain AI assistant with LLMs, agents, RAG, fine-tuning, LLMOps and AI systems techniques.

Distributed LLM inference. Connect home devices into a powerful cluster to accelerate LLM inference. More devices means faster inference.

Add a real-time analytics node to your operational database. Spice is a portable, accelerated SQL query, search, and LLM-inference engine in Rust for data-grounded AI apps and agents.

Open Source Pre-training Model Framework in PyTorch & Pre-trained Model Zoo

Open-source tools for prompt testing and experimentation, with support for both LLMs (e.g. OpenAI, LLaMA) and vector databases (e.g. Chroma, Weaviate, LanceDB).

Community maintained hardware plugin for vLLM on Huawei Ascend

High-performance inference framework for large language models, focusing on efficiency, flexibility, and availability.

LLM Finetuning with peft

A general fine-tuning kit geared toward image/video/audio diffusion models.

Paw Work - selection-first web agent for Chrome: select on the live page, describe the outcome, take away an editable office file. BYOK, sandboxed, no server.

AI生成の日本語を自然な日本語へ推敲するAgent Skill / Agent Skill for Refining AI-Generated Japanese into Natural Japanese

Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads

Global edition of ARTEX — an LLM multi-agent autonomous penetration-testing framework. English default, switchable to Korean/Chinese/Spanish/Japanese. For authorized, defensive and detection use only. (upstream: Autumn-27/ARTEX, AGPL-3.0)

🤖 Open-source, local-first alternative to Grok Bot, OpenAI dots and Claude Duet. Your own AI bots on any model: GPT, Claude, Gemini, Grok, DeepSeek, Kimi, GLM and 70+ more through one Modelflare key. Reach them in the app, WeChat or Feishu; they remember you, work in the background, run commands and use your computer. Pay per use.

streamline the fine-tuning process for multimodal models: PaliGemma 2, Florence-2, and Qwen2.5-VL

Build, personalize and control your own LLMs. From data pre-processing to fine-tuning, xTuring provides an easy way to personalize open-source LLMs. Join our discord community: https://discord.gg/TgHXuSJEk6

The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.