AI News

Sand.ai Completes $200 Million Series A+ Financing, Raising a Cumulative $300 Million in 10 Months

ZhidongxiUpdated Nov 6 03:04

AI summary

Shanghai-based video generation foundation model startup Sand.ai (Sanda Technology) announced the completion of a $200 million Series A+ financing round, bringing its cumulative funding to $300 million within 10 months; the company said that following this round it has officially joined the ranks of unicorns. This round was co-led by Cornerstone Capital and Wangsu Science & Technology, with participation from Shanghai state-owned-backed institutions including Pudong Venture Capital and Shanghai Guofang; the funds will be used for video foundation model R&D, training and inference infrastructure, product iteration, and global market expansion. In August 2026, the company open-sourced MAGI-2 Preview, which adopts an ultra-fine-grained MoE architecture with approximately 114B total parameters and about 6B activated per forward pass, and said that the official version of MAGI-2 with a new generation of larger parameter scale will be released soon.

Why read this

AI score 58

Beyond the financing details, the article also outlines Sand.ai's video model roadmap from autoregressive to hundred-billion-parameter MoE, which can be compared with its open-source cadence to trace its technical evolution.

Topics

Tags

Body · AI translation

Zhihui Dongxi reported on November 5 that today, Shanghai video generation foundation model startup Sand.ai (Sanda Technology) announced the completion of a $200 million (approximately RMB 1.334 billion) Series A+ funding round. This is the company's third funding round completed in 2026, bringing its cumulative funding to $300 million (approximately RMB 2.001 billion) within 10 months.

Sand.ai did not disclose its post-funding valuation. The company said that after this round of funding, it has officially joined the ranks of unicorns.

This funding round brought together industrial capital, well-known RMB and USD funds, Shanghai state-owned capital, and other investors. Sand.ai said the funds will mainly be used for video foundation model R&D, training and inference infrastructure construction, product iteration, and global market expansion. The official version of Sand.ai's next-generation, larger-parameter MoE architecture video generation foundation model MAGI-2 will be released soon.

This funding round was co-led by Cornerstone Capital and Wangsu Science & Technology; well-known institutions such as Times Capital, Panlin Capital, and Orient Securities Capital participated; Tianchuang Capital, together with Shandong Tietou Industrial Investment and Shandong Railway Fund, participated strategically; Shanghai state-owned background institutions such as Pudong Venture Capital and Shanghai Guofang joined; and existing shareholders such as Matrix Partners China, Huaye Tiancheng, and CAS Star continued to increase their investment.

Sand.ai was launched in January 2024 and is headquartered in Shanghai. Its business focuses on the R&D of video generation foundation models and application products.

The company's founder and CEO, Cao Yue, received a Ph.D. in Software Engineering from Tsinghua University in 2019. He previously served as a principal researcher at Microsoft Research Asia and head of the Vision Model Research Center at the Beijing Academy of Artificial Intelligence. He was also a co-founder of Light Year, an AI large-model company once acquired by Meituan. Cao Yue participated in research related to vision models such as Swin Transformer, Video Swin, SimMIM, and EVA, among which Swin Transformer won the ICCV 2021 Marr Prize.

Based on public information, Sand.ai has completed more than four funding rounds in total. In June this year, Sand.ai said it had completed two rounds of funding totaling more than $100 million, with investors including Look Capital, Lollapalooza Capital, JiuKun Venture Capital, Matrix Partners China, MSA Capital, Innovation Works, Xianghe Capital, Source Code Capital, CAS Star, Hongtai Fund, Capital Today, Huaye Tiancheng, Yunhui Capital, IDG, Baidu Ventures, and other leading institutions jointly investing.

▲Sand.ai's past funding situation (Image source: Qichacha)

Sand.ai continues to advance the evolution of video generation model architecture. Its technical route has expanded from autoregressive generation and joint audio-video modeling to hundred-billion-parameter-level MoE models, with optimization focused on model capabilities and inference efficiency.

In April 2025, Sand.ai released and open-sourced the autoregressive video generation model MAGI-1, which generates video clips in chronological order and supports continuous video generation and long-duration extension.

In March 2026, the company further open-sourced the audio-video foundation model MagiHuman, which adopts a unified single-stream Transformer architecture to jointly model text, visuals, and sound; and through infrastructure technologies such as MagiAttention and MagiCompiler, it improves the training and inference efficiency of large-scale video models.

In August 2026, Sand.ai released and open-sourced MAGI-2 Preview, becoming the world's first team to scale a video generation model to the hundred-billion-parameter level and open-source the model weights. The model adopts an ultra-fine-grained MoE architecture, with about 114B total parameters and only about 6B parameters activated per forward pass. On the image-to-video leaderboard of the third-party evaluation platform Artificial Analysis, MAGI-2 Preview ranked sixth globally at the time of its release.

In addition, Sand.ai also launched the video generation product VidMuse. Sand.ai said that two months after launch, the product's annual recurring revenue (ARR) exceeded $10 million.

After completing this funding round, Sand.ai will continue to advance the R&D of next-generation large-scale video models, accelerate the implementation of open-source foundation models and application products, and expand the global developer ecosystem and commercialization scenarios. Against the backdrop of intensifying competition in video generation models, whether the company can translate model capabilities into stable product revenue, developer ecosystem, and overseas market share will become the main focus of the next stage.

Source

Zhidongxi