Open source. Open possibilities.

Bob is a macOS menu bar translation and OCR tool, supporting selection, screenshot, and input translation, as well as screenshot OCR, silent recognition, QR code recognition, with integration of multiple translation and OCR services, offline recognition, and speech synthesis.

SGLang is a high-performance serving framework for large language models (LLMs) and multimodal models, designed for low-latency and high-throughput inference.

A local-first browser application that integrates AI chat, multi-window management, session isolation, a built-in terminal, and a CLI interface for controlling workflows from tools like Claude Code, Codex, and Gemini CLI.

Open-source AI-native call center: voice AI answers first over OpenAI Realtime or Qwen, hands off to human agents on FreeSWITCH queues with a browser softphone. One Go binary with REST API, SSE, embedded React UI, and PostgreSQL.

OmniRoute is an open-source MIT-licensed AI gateway providing one unified endpoint to access 352+ AI providers and 1200+ models (150+ free tiers). It features quota-aware auto-fallback, token compression saving 15–95%, and integrates with Claude Code, Cursor, Cline, and other coding agents.

FastGPT is a knowledge base platform based on large language models, providing out-of-the-box capabilities such as data processing, RAG retrieval, and visual AI workflow orchestration, supporting rapid construction and deployment of complex Q&A systems.

一个专为 Cloudflare Workers AI 设计的零成本网关。通过多账号免费额度叠加、智能路由和熔断机制，将 Qwen3 等模型转化为标准 OpenAI API，支持 Cherry Studio 等客户端接入。

MoziAI-35B is a locally deployable open-source multimodal LLM with a 35B MoE architecture compressed to ~15.9 GB via MoziSmartBit quantization, offering 256K context, vision, tool calling, and a finance-focused reasoning framework.

Open Interpreter is a terminal-based coding agent optimized for low-cost AI models, supporting multiple agent harnesses, ACP/Codex compatibility, sandboxed execution, and portable standards like AGENTS.md and MCP.

vLLM is a high-throughput, memory-efficient library designed for Large Language Model (LLM) inference and serving.

A growing collection of 61 computer vision tutorials covering state-of-the-art models like YOLO11, SAM 3, RF-DETR, and Qwen3-VL for object detection, segmentation, OCR, and more.

SupaNexus is an open AI gateway that unifies access to multiple LLM providers via one API, simplifying development with OpenAI-compatible interfaces and centralized model management.

Xinference is a unified inference library for serving language, speech, and multimodal models. It provides OpenAI-compatible RESTful API, heterogeneous GPU/CPU utilization, distributed deployment, and integrates with LangChain, LlamaIndex, Dify, and Chatbox.

UL-SMF is a KV cache compression framework for long-context Transformer inference, claiming up to 384x compression via FSQ and 16D latent mapping. The core compression logic requires a proprietary closed-source binary.

LlamaFactory is a Python framework for efficient fine-tuning of 100+ large language and multimodal models, offering zero-code CLI, a Gradio Web UI, LoRA/QLoRA, and multiple training algorithms.

Langchain-Chatchat is an open-source offline RAG and Agent application based on Langchain and ChatGLM, Qwen, Llama etc., supporting local knowledge base Q&A, multiple model inference frameworks, WebUI and API.

Qwen Code is an open-source AI coding agent for terminal, editor, desktop, browser, and chat. It supports multi-protocol APIs, local models, and dynamic agent workflows.

Ollama is a tool that allows users to run, manage, and build with open-source large language models locally on macOS, Windows, and Linux.

Hugging Face Transformers is a model-definition framework for machine learning across text, vision, audio, and multimodal tasks. It provides a unified API for inference and training with over 1 million pretrained checkpoints.