Aishi Technology

2mos agoupdate 2,038 0 0

Focusing on the field of AI video generation, with self-developed DiT architecture and multimodal generation technology as the core, to create a full-scene video generation solution from professional creation to mass entertainment.

Location:
Beijing China
Language:
zh
Collection time:
2024-07-17
爱诗科技Aishi Technology

Company Overview

AISI Technology isAI Video GenerationThe company is a leading company in the field, focusing on reducing the threshold of video creation and improving the efficiency of content production through self-developed technology. With the Diffusion+Transformer (DiT) architecture as the core, the company has built the world's leading video generation model PixVerse, whose latest version V5 supports 4K HD generation, second response, and ranked No. 1 in the world in the task of graphic video. The product matrix covers the overseas version of PixVerse and the domestic version of "Shoot Me AI", providing diversified functions such as text-generated video, graph-generated video, and character-driven video, which can satisfy the diversified needs from professional film and TV production to public entertainment.

Technologically, Aishi Technology has optimized the complex scene performance and instruction following ability through innovations such as spatio-temporal attention mechanism; ecologically, the open platform API has accessed more than 100 partners covering advertising, e-commerce, micro-short dramas and other fields. With a cumulative financing of more than 1 billion yuan and the support of Alibaba, Ant Group and other strategic investors, the company is accelerating the universalization of technology, promoting theAI Video GenerationTransition from a professional tool to a universal creation tool, reshaping the paradigm of content production and distribution.

Milestones

  1. Startups and Financing Breakthroughs
    • August 2023: Closed tens of millions of dollars in angel round financing to initiate technology pre-research.
    • 2024::
      • February: Multi-modal generation algorithm through the record, support text-generated map, text-generated video, map-generated video function.
      • April: Received over 100 million yuan of A2 round of financing led by Ant Group, with cumulative financing of over 200 million yuan.
      • JulyPixVerse V2, which adopts the Diffusion+Transformer (DiT) architecture, has been released, introducing a self-developed spatio-temporal attention mechanism to enhance the performance capability of complex scenes.
      • December: Completed rounds A2 to A4 of financing totaling nearly $300 million.
    • 2025::
      • March: Completed A5 round of financing, with cumulative financing of over 400 million RMB.
      • September 10: announced the completion of more than 60 million U.S. dollars in Series B financing, led by Alibaba, Brilliance Caixin, Shenzhen Venture Capital and other institutions to follow the investment, the largest single financing record in the field of domestic video generation.
      • October 17: Completed RMB 100 million B+ round of financing, co-invested by Fosun Ruizheng, Tongchuang Weiye and Shunxi Fund.
  2. Product Iteration and Market Expansion
    • January 2024: Launched an overseas version of its product, PixVerse, which focuses on template-driven video generation and has quickly taken the world by storm.
    • June 2025: Launched "Shoot Me AI" for the domestic market to meet the needs of localized creation.
    • August 2025: The open platform API has accumulated more than 10 million times of video generation and more than 100 partners.

core product

  1. PixVerse
    • functionality: Support text-generated video, graph-generated video, 4K resolution generation, V5 version on line in August 2025 to achieve second generation, optimize dynamic effects, ultra-high-definition visual processing and command following capabilities.
    • Technical Highlights::
      • The world's first video generation model based on DiT architecture, which has undergone five rounds of iterations and eight version updates.
      • In Artificial Analysis, a third-party measurement platform, the Tusheng video task was ranked No. 1 globally, and the Wusheng video task was ranked No. 2.
    • market performanceThe company was ranked 25th on a16z's "Global Top 50 Generative AI Consumer Mobile Apps" list, with a monthly increase of over 26% in visits to its website.
  2. Shoot Me AI
    • localizationThe platform is a lightweight AI video generation platform for domestic users, emphasizing "fast speed, easy to start, and controllable creativity".
    • functionality: Provide creative templates such as character-driven videos, first and last frame sequels, and 3D artifact generation to lower the threshold of creation.
    • ecological construction: Deep integration with partners through API interfaces to support advertising, film and TV, game production and other scenarios.

development prospect

  1. Technology Inclusion
    • Through the "free + membership" model to reduce the threshold of creation, to promote the AI video generation from professional creation to public entertainment penetration.
    • The expansion of the open platform API and partner ecology will accelerate the landing of the technology in advertising, film and television, e-commerce and other scenes.
  2. Vertical Scene Deepening
    • Standardized solutions are formed in the fields of e-commerce marketing and micro-short dramas, such as generating content such as product display videos and virtual anchors through AI.
    • Combined with the advantages of short-video platform dissemination, it has spawned explosive content with over 10 million plays.
  3. Policy and Compliance Advantage
    • Relying on the "14th Five-Year Plan for the Development of Film Science and Technology" and other policy support, we have established a unique advantage in the areas of virtual production and content auditing.
    • Addressing compliance challenges in commercial deployments through the construction of data security and intellectual property protection systems.
  4. Globalization and competitive differentiation
    • Differentiated from overseas giants (e.g. OpenAI Sora), China's "Cost Reduction and Efficiency" focuses on reducing costs through architectural innovations such as the Mixed Expert System (MoE), which is better adapted to the needs of small and medium-sized enterprises (SMEs).
    • In terms of business model, China adopts "free trial + traffic sharing" to attract users, and combines with short video platform to realize the efficient dissemination of content.

data statistics

Relevant Navigation

No comments

none
No comments...