FinSight: Towards Real-World Financial Deep Research. 🎯One ticker, one click, one publication-ready report.

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

An Open Phone Agent Model & Framework. Unlocking the AI Phone for Everyone

“AI-Compass”将为社区指引在 AI 技术海洋中航行的方向，无论你是初学者还是进阶开发者，都能在这里找到通往 AI 各大方向的路径。旨在帮助开发者系统性地了解 AI 的核心概念、主流技术、前沿趋势，并通过实践掌握从理论到落地的全过程。

Scalable toolkit for efficient model reinforcement

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

PyTorch compiler that accelerates training and inference. Get built-in optimizations for performance, memory, parallelism, and easily write your own.

All-in-one training for vision models (YOLO, ViTs, RT-DETR, DINOv3): pretraining, fine-tuning, distillation.

A framework for efficient model inference with omni-modality models

LLM inference in C/C++

Open Vision Agents by Stream. Build Vision Agents quickly with any model or video provider. Uses Stream's edge network for ultra-low latency.

Easy Data Preparation with latest LLMs-based Operators and Pipelines.

Multi-agent framework, runtime and control plane. Built for speed, privacy, and scale.

Tutel MoE: Optimized Mixture-of-Experts Library, Support GptOss/DeepSeek/Kimi-K2/Qwen3 using FP8/NVFP4/MXFP4

Extended LaTeX template for CVPR/ICCV papers

A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations

PyTorch-native post-training at scale

A simple, any-to-any modality framework for pretraining and finetuning. Lean, flexible, and built for research.

A JAX-native LLM Post-Training Library

Youtu-Embedding is an industry-leading, general-purpose text representation model developed by Tencent Youtu Lab.

Trae Agent is an LLM-based agent for general purpose software engineering tasks.

Official repository of Youtu-GraphRAG: Vertically Unified Agents for Graph Retrieval-Augmented Complex Reasoning

Official Code for "Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Search"

A simple yet powerful agent framework that delivers with open-source models