Chinese AI · 71 reviewed
China ships more open-weight models than anywhere else and almost none of it is documented properly in English. These are reviewed from Shenzhen, with the licence terms read in full, the mainland and international endpoints tested separately and the practical catches noted: which weights are on Hugging Face rather than ModelScope only, which APIs need a mainland phone number and which licences change between releases.
Managed vector database built on Milvus, the open-source engine behind most billion-scale retrieval systems, for teams who need hybrid search without running the cluster themselves.
ByteDance's AI IDE and standalone SOLO agent that scaffolds projects, runs terminal commands and tests in a browser, at $10 a month for Pro against Cursor's $20.
Unitree's open robot learning stack: world models, a vision-language-action policy, Isaac and MuJoCo training environments and the SDK that drives Go2, G1 and R1 hardware.
Open-source LLM app platform from LangGenius for agentic workflows and RAG pipelines. Modified Apache 2.0: self-host freely, multi-tenant SaaS needs a commercial licence.
InfiniFlow's Apache 2.0 RAG engine built on deep document parsing. DeepDoc handles layout, tables and OCR before chunking, and every citation points back to a highlighted region in the source PDF.
OpenBMB's tokenizer-free TTS, Apache 2.0 on weights and code. VoxCPM2 is 2B parameters, 30 languages, 48kHz output and voice design from a written description with no reference clip.
Alibaba Qwen team's 20B MMDiT image foundation model plus its Edit and Layered variants, all Apache 2.0. Renders long-form Chinese and English text inside images at commercial quality.
Shanghai AI Lab's open vision-language family (书生·万象), nine sizes from 1B to 241B-A28B under Apache 2.0. The default open VLM for OCR, document parsing and GUI agent work.
MiniMax's open-weight flagship LLM (稀宇科技), distinct from the Hailuo video product: 428B mixture-of-experts, 23B active per token, 1M context and native image and video input.
Ant Group's open-weight MoE family (百灵, Bailing): Ling-3.0-flash at 124B total and 5.1B active, a 7.9B tiny variant and trillion-parameter Ring reasoning models, all MIT.
OpenBMB's on-device line (面壁小钢炮): MiniCPM5-1B at 1.08B parameters with a 131K context, plus MiniCPM-V 4.6 multimodal at 1.3B. Apache 2.0 on weights and code.
Apache 2.0 vector database from Zilliz, self-hostable from a pip install up to a billion-vector cluster. The open engine under Zilliz Cloud, minus the proprietary Cardinal search core.
MPP analytics database that began as Baidu's Palo. Version 4.0 put HNSW vector indexes, BM25 full-text search and SQL-callable LLM functions into the same engine that runs the OLAP queries.
Open manipulation data and policy weights from AgiBot (智元机器人): a million-plus real robot trajectories, the GO-1 VLA and the Genie Envisioner world model, all non-commercial.
BAAI's embodied brain model (悟界·RoboBrain), Apache 2.0 at 4B and 8B, shipped in parallel NVIDIA and Moore Threads builds. It plans and points in 3D for a controller underneath.
Baidu's Apache 2.0 autonomous driving platform (阿波罗). Version 11.0, January 2026, aims squarely at sweepers, delivery pods, patrol vehicles and campus shuttles rather than robotaxis.
MIT-licensed official CLI for Lark and Feishu (飞书), from ByteDance. 200+ commands and 24 agent skills over Docs, Base, Sheets, Calendar and approvals, built for AI agents.
Kingsoft's enterprise office AI platform (金山办公 WPS 365). AI runs inside the document, and Qingzhou AI puts private deployment on a single CPU server. From 199 RMB per seat a year.
MIT-licensed open agent framework from the MetaGPT team, prototyped in three hours as the open answer to Manus. 58,100 stars and still pushing commits in August 2026.
Image and text to 3D from VAST in Beijing. Hosted Tripo 3.0 Ultra reaches 2 million polygons, while TripoSG 1.5B, TripoSR and UniRig ship as MIT-licensed weights you can run yourself.
Apache 2.0 LLM serving and quantization toolkit from Shanghai AI Laboratory's InternLM group. Builds for NVIDIA, Huawei Ascend, Cambricon and AMD ROCm from a single codebase.
Apache 2.0 agent framework from Alibaba Tongyi Lab's SysML team. Version 2.0 folded the separate Runtime project into the core, so sandboxing, agent serving and observability now ship as one...
Baidu's Apache 2.0 OCR toolkit (飞桨 PaddleOCR). PP-OCRv6 ships at 1.5M, 7.7M and 34.5M parameters, and PaddleOCR-VL-1.6 is a 0.9B document parser covering 109 languages.
RedNote's (小红书) MIT-licensed document parser. A single vision-language model does layout, reading order and text; dots.mocr at 3B added direct SVG output for charts in March 2026.
Alibaba's FunAudioLLM TTS line, Apache 2.0 on both code and weights. Fun-CosyVoice3-0.5B covers 9 languages and 18+ Chinese dialects with roughly 150ms bi-streaming latency.
Bilibili's zero-shot TTS with the finest emotion and duration control in the open field. IndexTTS-2.5 is 0.8B across five languages, but commercial use needs written bilibili authorization.
Alibaba Tongyi Lab's open-weight video generation family (通义万相 Wan). Apache 2.0 mixture-of-experts checkpoints you can download and fine-tune, separate from the closed Wan 3.0 API.
Ant Group's OceanBase team open-sourced this AI-native search database under Apache 2.0 (seekdb, from the OceanBase 蚂蚁 stack). Vector, full-text and scalar filters in one SQL query, embeddable in...
PingCAP's MySQL-compatible distributed SQL database, Apache 2.0 across TiDB, TiKV and TiFlash. Vector columns and HNSW indexes ride on the TiFlash columnar replica, next to transactional rows.
Alibaba's open-weights coding model family plus the Apache 2.0 Qwen Code terminal agent, separate artifacts from the hosted Qwen Chat product. Qwen3-Coder-Next runs 80B total, 3B active.
Ant Group's open-source code intelligence stack (蚂蚁 CodeFuse): an Apache 2.0 GödelScript static analysis engine, MIT-licensed Ling-Coder-Lite weights and an OpenSumi-based AI IDE.
The open repos behind Unitree's robots (宇树科技): BSD-3 C++ and Python SDKs, MuJoCo and IsaacLab RL training, ROS 2 packages and XR teleoperation for data capture.
Alibaba's open model hub (魔搭社区). Apache 2.0 client library, 170,000+ open-weight models and 2,000 free API calls a day. Where most Chinese labs publish first.
Huawei Cloud's full-lifecycle AI platform. Training and serving on Ascend NPUs through CANN, with Lite Server, Lite Cluster and a MaaS layer. Proprietary and quote-priced.
Tencent's consumer assistant app (腾讯元宝), distinct from the Hunyuan model family it serves. Its real edge is legal read access to the closed WeChat corpus. Mainland only.
Design agent from LiblibAI (哩布哩布) that plans an entire deliverable set from one brief then builds it on an editable canvas. Ships as 星流 Agent inside mainland China.
Tencent's zero-dependency C++ inference library (腾讯 ncnn), BSD 3-Clause. Runs inside WeChat and QQ, with prebuilt libraries for HarmonyOS, WebAssembly, RISC-V and LoongArch.
Baidu's training and inference framework (飞桨), Apache 2.0, with kernels for Kunlunxin XPU, Ascend NPU, Hygon DCU and Cambricon MLU alongside CUDA. PaddleOCR is why most teams arrive.
Huawei's Apache 2.0 framework (昇思) built around Ascend NPUs. Development is on Gitee, the GitHub mirror is stuck on 2024 and MindSpore Lite is the built-in AI engine in HarmonyOS.
Document-to-Markdown engine from Shanghai AI Lab's OpenDataLab. Relicensed off AGPLv3 in April 2026, reported at 95.4 on OmniDocBench v1.6 and running on ten domestic AI chip families.
PaddleOCR's models converted to ONNX and shipped without the PaddlePaddle framework. Apache 2.0, a 27.3 MB wheel, PP-OCRv6 small by default and fully offline CPU deployment.
Alibaba DAMO's industrial ASR toolkit. MIT code, Apache 2.0 Fun-ASR-Nano weights, 1.80% Chinese CER, plus VAD, punctuation, diarization and an OpenAI-compatible self-hosted server.
Tencent's open image-to-3D model with PBR texture synthesis and full training code. A separate artifact from the Hunyuan chat product, and not under an OSI-approved licence.
Shanghai AI Lab's open-weight family (书生·浦语), now the Intern-S scientific multimodal line: 35B up to 1T parameters, Apache 2.0, trained on domestic computing infrastructure.
Alibaba's agentic IDE, which absorbed Tongyi Lingma (通义灵码) as Qoder CN on 20 May 2026. Repo Wiki and Quest Mode on credit-based pricing, with Linux builds since version 1.0.
Tencent Cloud's coding assistant (腾讯云代码助手), a distinct product from the Hunyuan chatbot. Standalone IDE, editor plugins and a CLI with native Agent Client Protocol support.
General AI agent from Butterfly Effect, the Wuhan and Beijing team now based in Singapore. Runs long multi-step tasks in a cloud VM. Blocked from mainland China since 2025.
Deemos (影眸科技), a ShanghaiTech spinout, builds Rodin. Gen-2.5 emits native quad meshes at 4k to 50k quads, which is why it survives real pipelines. Hosted only, no weights.
Manycore Tech (群核科技) of Hangzhou, behind Kujiale and Coohom, open-sourced its indoor spatial stack. SpatialLM turns point clouds into structured layouts, SpatialGen generates scenes.
Beijing engineering team, Santa Clara registration, founded by Taichi author Yuanming Hu. Meshy 6 reached general availability on 18 January 2026 with meshes to roughly 600K faces.
Alibaba's on-device inference engine (阿里 MNN), Apache 2.0. Runs Qwen3, Qwen3-VL and DeepSeek distills on phones across CPU, OpenCL, Metal, Vulkan, CUDA and Qualcomm Hexagon.
Chinese AI lab producing open-source LLMs that rival proprietary models at a fraction of the cost, with strong reasoning and coding abilities.
Baidu’s AI image generation platform offering text-to-image creation with strong Chinese artistic style generation and commercial design capabilities.
ByteDance’s AI image and video generation platform integrated with Jianying (CapCut’s Chinese version), offering creative content generation tools.
Kunlun Tech’s AI search and assistant platform offering real-time web search with AI-powered answers and content generation.
Alibaba’s AI-powered search and productivity app combining intelligent search, document tools, cloud storage, and AI assistant capabilities.
Meitu’s suite of AI-powered image and video editing tools, specializing in portrait enhancement, commercial photography, and design automation.
Alibaba’s AI image generation platform offering text-to-image creation with strong aesthetic quality and Chinese cultural understanding.
SenseTime’s comprehensive AI platform offering large language models, image generation, video generation, and enterprise AI solutions.
Kai-Fu Lee’s open-source AI model family offering high-performance bilingual models with strong benchmark results across reasoning and coding tasks.
Baichuan AI’s open-source large language models focused on Chinese language understanding and enterprise AI applications.
Tencent’s large language model and AI platform, integrated across WeChat, Tencent Cloud, and the broader Tencent ecosystem.
Flytek’s AI assistant powered by the Spark cognitive model, with industry-leading Chinese speech recognition and voice AI capabilities.
Kuaishou’s AI video and image generation platform that has gained global recognition for producing high-quality, realistic AI videos from text and image prompts.
MiniMax’s consumer AI platform offering conversational AI, industry-leading AI video generation, and music creation capabilities.
hipu AI’s bilingual AI assistant powered by the GLM-4 model series, with strong academic roots from Tsinghua University and excellent bilingual capabilities
Moonshot AI’s long-context AI assistant, pioneering ultra-long document processing with a 2 million token context window.
ByteDance’s AI bot-building platform that enables anyone to create custom AI assistants with plugins, workflows, and knowledge bases without coding.
ByteDance’s AI chatbot and productivity assistant, one of China’s most popular AI apps with deep integration across ByteDance’s ecosystem including Feishu (Lark).
Alibaba’s open-source large language model family and AI assistant, offering strong multilingual capabilities and deep enterprise integration with Alibaba Cloud.
Baidu’s flagship AI chatbot powered by ERNIE 4.5 and reasoning model ERNIE X1, offering multimodal capabilities across text, image, audio, and video.
Licences on Chinese model releases change more often than most teams expect, sometimes between minor versions of the same family. Every entry here records the licence verified at the time of review. Check the LICENSE file on the release you actually pull.