Skip to main content

Evoken

Category: Generative Media (Image / Video)

Parent company of LiblibAI, a Chinese AI creative content technology group building a multi-product matrix spanning AI image generation (LiblibAI), AI video creation (LibTV), and AI design agent (Xingliu/Lovart). Evoken was founded in 2023. The company is led by Chen Mian (陈冕). Based in Beijing, China. Team size: 101-500. Total funding raised: $470.0M. Latest round: Series B. Key investors include Granite Asia, Tencent, Shunwei Capital, Hongshan Capital (Sequoia China), Ant Group, Gaorong Capital, Source Code Capital, Mingshi Venture Capital, CMC Capital, GSR Ventures, Yingce Capital, HT Investment.

Founded
2023
Headquarters
Beijing, China
Team size
101-500
Total funding
$470.0M

Value proposition

China's first full-stack AI creative content ecosystem spanning image, video, and design — enabling creators to complete the entire workflow from inspiration to finished product delivery within a single platform ecosystem.

Products and solutions

1) LiblibAI (哩布哩布AI) — AI image generation platform & creator community, 30M+ users, 500K+ original models, China's largest AI material website, 2) LibTV — AI video creation platform launched March 2026, serves short drama teams, film/TV production, ad agencies, 3) Xingliu (星流) / Lovart — AI design agent launched July 2025, full-stack intelligent design capabilities, 4) Lovart (overseas brand) — world's first design vertical agent product

Unique value

Only Chinese AI application company with a complete product matrix (community + video + design) that has achieved multi-product PMF, generating $300M ARR with 30x YoY growth. First AI video tool designed for both human creators and AI agents simultaneously.

Target customer

Content creators, designers, short drama production teams, film/TV production studios, advertising agencies, brand clients, individual creators

Industries served

AI creative content, AI image generation, AI video production, AI design, short drama production, advertising, film/TV production, brand marketing

Technology advantage

Multi-product PMF validation; proprietary "system-first" architecture for AI video (LibTV); integration of multiple multimodal models (GPT-4o, Gemini, Claude, Seedance, Kling, etc.); ChatCanvas interactive tool for real-time design modification via natural language; deep understanding of content creation workflows productized into AI tools; strong creator ecosystem with network effects (30M+ users, 500K+ models)

How they differentiate

Unlike most AI application companies that focus on single products, Evoken has built a diversified product matrix (LiblibAI for community/assets, LibTV for video production, Xingliu for design delivery) that creates a self-reinforcing ecosystem. First Chinese AI app company to achieve "group-ization" strategy. LibTV is the first video tool designed from day one for both human creators and AI agents. Revenue comes from three synergistic products rather than a single source.

Main competitors

Baidu Wenxin Yige (文心一格), ByteDance Jimeng (即梦), Alibaba Tongyi Wanxiang (通义万相)

Key partnerships

Alibaba Cloud (full-stack AI capabilities for global expansion), Tencent (Hunyuan Image 3.0 model integration), Step-Video-T2V model integration, Alibaba Tongyi series model integration, Kling 3.0 video model integration

Notable customers

Kingsoft Office, Wondershare Technology, G-bits Games, Tmall Campus, Tsinghua University, 1000+ short drama teams, film/TV production studios, advertising agencies, and brand clients

Major milestones

2023-05: Founded by Chen Mian (ex-ByteDance CapCut head), 2023-07: Angel round ($3.5M) from GSR, Gaorong, Source Code, 2024-07: Series A led by Mingshi Venture Capital, 2025-02: Series A+ from Shunwei, Yingce, 2025-07: Launched Xingliu Agent (AI design) and Lovart overseas, 2025-10: $130M Series B (largest AI app round in China in 2025), 2026-03: Launched LibTV (AI video creation), 2026-06: $300M Series B+ at $2B+ valuation, ARR reaches $300M, rebranded as Evoken group brand

Growth metrics

ARR: ~$300M (May 2026); 30M+ cumulative users (LiblibAI); 10M+ users (Xingliu); 500K+ original models; 100M+ professional images/videos; 3000% YoY revenue growth (May 2026); LibTV single-day revenue exceeded $1M in launch month

Market positioning

China's leading AI application company by funding and revenue. Positioned alongside global peers like Cursor, Suno, HeyGen, and Higgsfield in the first tier of AI application companies globally, but at a significantly lower valuation ($2B vs global peers).

Geographic focus

China (primary), expanding globally via Lovart brand (US and 70+ countries)

Patents and IP

Self-developed image generation model Star-3; proprietary "system-first" architecture for AI video creation

About Chen Mian (陈冕)

Ex-ByteDance Global Commercialization Head for CapCut/Jianying (video-editing app); previously Product Director at Mobike; also worked at Didi and Missfresh. Graduated from Southeast University (bachelor's). At 28, became the youngest Product 4-1 (senior management) in ByteDance history.

Official website: