Skip to main content

Rabbitpre Tech

Category: Foundation Models / LLMs

Chinese visual AI company building a self-developed visual-spatial intelligence large model (Tuling/兔灵) and enterprise AI marketing/interaction platform, including the SkillsUI agent middleware layer that orchestrates legacy enterprise systems via natural language. Rabbitpre Tech was founded in 2014. The company is led by 董少灵 (Shaoling Dong). Based in Shenzhen, Guangdong, China. Team size: 201-500. Total funding raised: $100.0M. Latest round: Series F. Key investors include 深创投 (SCGC); 腾讯投资 (Tencent Investment); 招商局创投 (China Merchants Venture); 招商资本 (China Merchants Capital); 越秀产业基金 (Yuexiu Industrial Fund); 第四范式 (4Paradigm); 南山战新投; 润信新观象; 嘉道; 中国风投; 青岛华通; 青岛海诺; 朗玛峰创投; 碚曦投资 (Withinlink); 基美影业; 龙岗金控; 晶泰科技 (XtalPi).

Founded
2014
Headquarters
Shenzhen, Guangdong, China
Team size
201-500
Total funding
$100.0M

Value proposition

Enterprise AI interaction platform that connects legacy business systems via an agent middleware layer (SkillsUI), enabling natural language-driven task execution across ERP, OA, CRM without replacing existing IT infrastructure.

Products and solutions

SkillsUI (native open AI interaction platform - agent middleware layer); AnyReal (photorealistic product image generation); BOOM (AI video production platform); DeiNai (overseas AI marketing platform); Tuling/兔灵 visual-spatial intelligence large model; Open-Sora Plan (open-source video generation); UniWorld-V1/V2 (unified understanding-generation architecture); RabbitPre Marketing Cloud (CMS, SCRM, CDP, MA)

Unique value

Only Chinese company with both a self-approved visual-spatial intelligence large model (Tuling - first in Guangdong province to pass备案) AND a production-grade agent middleware layer (SkillsUI) that turns legacy enterprise systems into AI-callable capabilities without system replacement.

Target customer

Large enterprises and mid-market companies in finance, insurance, retail, FMCG, real estate, healthcare, and government sectors needing AI-powered marketing and enterprise system orchestration.

Industries served

Financial services; Insurance; Retail/E-commerce; FMCG; Real estate; Healthcare; Government/Public services; Cross-border e-commerce

Technology advantage

Self-developed visual-spatial intelligence large model (Tuling); Open-Sora Plan (open-source video generation, #1 on GitHub trending); UniWorld-V2 (unified understanding-generation architecture outperforming OpenAI GPT-Image-1); MoE++ accepted at ICLR 2025 Oral; Video LLaVA used as benchmark in Google Gemini Pro technical report; First 100% Huawei Ascend-based visual generation model (Open-Sora Plan V1.5); Dragon Diffusion, Dragon Adapter, FreeDoM core technologies

How they differentiate

1) Self-developed visual-spatial intelligence large model (Tuling) - first in Guangdong to pass备案; 2) Open-Sora Plan open-source video model that topped GitHub global trending; 3) UniWorld-V2 model outperforming OpenAI's GPT-Image-1 on benchmarks; 4) SkillsUI agent middleware layer that doesn't replace legacy systems but adds an AI orchestration layer; 5) Full-stack from foundation model to enterprise application; 6) Deep partnership with Huawei Ascend (first large-scale user of Ascend 910C)

Main competitors

特赞 (Tezign) - AI content/marketing platform; 第四范式 (4Paradigm) - enterprise AI platform; 百度营销 (Baidu Marketing) / 阿里妈妈 (Alimama) - AI marketing platforms

Key partnerships

Huawei (Ascend 910C deep partnership, first large-scale user); Peking University (AIGC Joint Lab); Tencent (AI Accelerator, cloud partnership); Microsoft (Shanghai Accelerator); Alibaba Cloud; China Mobile; 第四范式 (4Paradigm); 招商局创投 (China Merchants Group VC); 深创投 (SCGC)

Notable customers

Huawei; Honor; 赛维时代 (Saiwei Times); 伊利 (Yili); 中国银行 (Bank of China); 平安集团 (Ping An Group); 深圳政数局 (Shenzhen Government Data Bureau); 中兴 (ZTE); 华润 (China Resources); 金地 (Gemdale)

Major milestones

2014: Founded in PKU dormitory; 2015: A轮融资 from 深创投; 2017: Forbes 30 Under 30 for founder; 2018: Selected for Microsoft Accelerator & Tencent AI Accelerator; 2021: C2轮融资 with Tencent, 招商局, 第四范式; 2023: D+轮融资; PKU-RabbitPre AIGC Joint Lab established; 2024: Open-Sora Plan #1 on GitHub global trending; National Specialized & New "Little Giant" Enterprise; 2025: E轮融资; UniWorld-V2 outperforms OpenAI GPT-Image-1; Guangdong first备案 visual model; AI commercialization annual contract value >¥100M; 2026: F轮融资; SkillsUI launch; IPO preparation (Hong Kong listing planned); Guangdong first AI National High-Skill Talent Training Base

Growth metrics

4100万+ enterprise users; 48万+ paid users; 300+ head enterprise customers; 100+ industry categories; 2025 AI commercialization annual contract value >¥100M (¥1亿); Monthly compound growth rate >60% for overseas business; Open-Sora Plan single model downloads >26M

Market positioning

Leading Chinese enterprise AI marketing and interaction platform with proprietary visual-spatial intelligence foundation model. Positioned as the "AI entry layer" for enterprise digital transformation, competing with both marketing cloud platforms and AI agent middleware providers. One of only two visual foundation model companies in China.

Geographic focus

China (primary), expanding to overseas markets via DeiNai cross-border marketing platform

Patents and IP

National High-Tech Enterprise; National-level Specialized and New "Little Giant" Enterprise; Guangdong Province first AI National High-Skill Talent Training Base; Multiple SOTA breakthroughs in visual understanding, compression reconstruction, backbone networks, conditional control

About 董少灵 (Shaoling Dong)

Peking University (PKU) - JSD (Doctor of Juridical Science) & International Law Master; Sun Yat-sen University - Management background. Forbes 30 Under 30 (2017). Founded RabbitPre in his PKU dormitory in 2014.

Official website: