Image source: Generated by Unbounded AI

It’s the end of the year, and suddenly there’s a lot of good news about AI company financing. Some people find it reassuring, but others think this is called a promotion game and encourage everyone to speak out more.

After all, many of my friends have raised money and are very low-key.

We have taken stock of the AI ​​financing cases that have been publicly disclosed in recent weeks.

Pixel Bloom/AiPPT: Round B2

Pixel Bloom (AiPPT.com), a well-known company in the AI ​​office track, announced the completion of Series B2 financing. This round of financing was led by the Beijing Artificial Intelligence Industry Investment Fund (managed by Shunxi Fund and Qiming Venture Partners), with participation from well-known institutions and star entrepreneurs.

This is the fifth round of financing that “Pixel Bloom” has received in four years. Previously, "Pixel Bloom" has received funding from A-share listed companies Vision China, Starlink Capital (Zhipu Ecological Z Fund), 36 Krypton, Xinyuan Capital, Weimeng Media, Xintian Venture Capital, Ceyuan Venture Capital, Asia Jie Fund, Shence Data, Xiaomi co-founder Wang Chuan and other institutions and star entrepreneurs have invested. This round of financing will be used for global market expansion, product line extension, AI capability improvement, etc.

Its products include AiPPT.cn (domestic)/AiPPT.com (overseas), AiDesign AIGC content center, 365 editor, AiDesign online design tool, AiH5, Dingdang Haoji Readlecture.cn, There are more than 10 content creation tools that apply AI capabilities, including Visdoc.Ai.

HiDream: Hundreds of millions of yuan in Series A

HiDream.ai, an AI video generation start-up, has completed a new round of financing after completing the Pre-A round of financing led by Dunhong Capital. Obtained Series A financing led by state-owned funds led by Hefei Industrial Investment, with a total financing scale of hundreds of millions of yuan. Also participating in the investment include Anhui Provincial Artificial Intelligence Fund of Funds and Hubei Provincial Yangtze River Film Group Co., Ltd.

The multi-modal large model "Zhixiang Big Model" self-developed by Zhixiang Future will be launched in August 2023, and the upgraded version of "Zhixiang Big Model 2.0" will be released in July this year. Zhixiang Future calls the "Zhixiang Big Model" the world's first open-to-use image and video generation Diffusion Transformer (DiT) architecture model, which realizes the interoperability of four modalities: text, image, video, and 3D. It can provide minute-level marketing video generation services for corporate customers, while providing 5- to 15-second video generation tools for individual users.

Aishi Technology: nearly 300 million yuan in Series A

Aishi Technology recently officially completed A2-A4 rounds of financing, totaling nearly 300 million yuan. Investors include Ant Group, Beijing Artificial Intelligence Industry Investment Fund, Guoke Investment and Light Source Capital.

When the core AI video generation product PixVerse was first launched in January 2024, PixVerse achieved more than 1.2 million visits in the first month after its launch. One year later, this number has been refreshed: PixVerse has more than 12 million global users, with nearly 6 million monthly active users. Moreover, the team has now achieved large-scale revenue.

Step Star: Hundreds of millions of dollars in Series B

Step Star has recently completed Series B financing, with the total financing amount reaching hundreds of millions of dollars. This financing involves state-owned assets, strategic and financial investors, etc. The core investors include Shanghai State-owned Capital Investment Co., Ltd. and its funds. Strategic and financial investors include Tencent Investment, Wuyuan Capital, Qiming Venture Capital, etc.

It is understood that this financing will be used to continue investing in basic model research and development, strengthen multi-modal and complex reasoning capabilities, and increase coverage of C-side application scenarios through products and ecology to provide a rich user experience.

From the perspective of the base model, Step Star is one of the few domestic models that has formed a full range of model matrices from understanding to generation, from text to multi-modal.

In the past 10 months, Stepstar has released a total of 11 self-developed base models, including the 100-billion-parameter language model Step-1, the trillion-parameter language model Step-2, and the image understanding model. Step-1V, large image generation model Step-1X, large multi-modal understanding model Step-1.5V, large video generation model Step-Video, large speech reproduction and generation model Step-tts-mini, large speech recognition model Step-Asr, and the recently released Step-1o Audio, China’s first end-to-end speech model with 100 billion parameters. Among them, Step-2 is the first trillion-parameter model released by a large model startup company in China.

Zhipu AI: 3 billion yuan

On December 17, Zhipu AI announced that it had recently completed a new round of financing of 3 billion yuan. New investors include a number of strategic investors and state-owned assets, as well as old shareholders such as Legend Capital. Continue to invest.

This is the second round of financing recently available for Zhipu AI. In September 2024, Zhongguancun Science City Company announced that it would lead the investment in Zhipu with a pre-investment valuation of 20 billion to support the technological innovation and ecological development of domestic large-scale base models.

This also makes Zhipu AI not only one of the first large-model startups with a valuation exceeding 20 billion, but also one of the highest-valued large-model startups in China.

Zhipu AI has currently released the video generation model CogVideoX that benchmarks against Sora, and the end-to-end voice models GLM-4-Voice and GLM-4-VideoCall that benchmark against GPT-4o, taking the lead in launching independent AutoGLM, an intelligent agent that operates a mobile phone.

Wall-Facing Intelligence: Hundreds of Millions

Wall-Facing Intelligence completed a new round of financing of hundreds of millions of yuan. This round of financing was jointly led by Loongson Venture Capital, CDH Baifu, Zhongguancun Science City Fund and SAIF Fund. Beijing Artificial Intelligence Industry Investment Fund and Zero2IPO Venture Capital followed in the investment, and Wanjia Capital served as the exclusive financial advisor for this round. Its last round of financing was announced in April 2024, and the amount was also hundreds of millions of yuan in financing.

After the completion of this round of financing, Wall-Facing Intelligence will further accelerate the commercial layout of efficient large-scale models represented by end-side AI, with the same parameters, higher performance, lower energy consumption, and faster speed. The model deeply serves the industry. Since its release in February this year, the cumulative downloads of the MiniCPM series have exceeded 3 million, and it has repeatedly topped the trend lists of GitHub and HuggingFace large models, becoming a world-renowned star project in the open source community.

Dream Dimension: tens of millions in Series A

AI content platform Dream Dimension recently completed tens of millions of yuan in Series A financing. This round of financing was led by Jinqiu Fund, Mingshi Venture Capital, Gaorong Venture Capital, SenseTime and many other well-known institutions and major AI companies participated in the investment. This financing will provide strong support for the company to accelerate technology research and development and market expansion, and help the company's innovative exploration in the fields of generative AI and interactive content.

Dream Dimension was founded in 2023 and focuses on a new generation of content platform driven by multi-modal AI. The company's core product "Dream Dimension" is positioned to serve young users and combines generative AI technology to provide an interactive and dramatic virtual character interaction experience.

In the field of AI content consumption, Dream Dimension creates high-quality and interactive multi-node content through the UGC model, covering a variety of themes such as campus, ancient style, time travel, science fiction, etc., and can also adapt to reality The diversified content such as themes, fan creations, and memes shows a strong ability to generalize content.

Shanji: Series A exceeds 100 million yuan

On December 18, smart hardware brand "Shanji Technology" once again announced that it has received Series A+ financing exclusively invested by Oasis Capital, with Yuewei Capital serving as the exclusive financial advisor. So far, within one month, Shanji has completed a series A financing of over 100 million yuan. The previous Series A financing was led by Guangyuan Investment, followed by Future Light Cone Frontier Technology Fund and edge AI chip company Yuntian Lifei.

On December 19, Shanji launched its first AI glasses - Shanji AI "Paipai Mirror" A1. This product is based on the design blueprint of "making a good pair of glasses" and cooperates with manufacturers in various fields such as Sony Semiconductor, AAC AAC Technology, LOHO, and Yuntian Lifei. Shanji AI camera supports access to dozens of large models, including Wenxinyiyan, iFlytek Spark, Byte Doubao, Tongyi Qianwen, Tencent Hunyuan, Kimi, Zhipu, SenseTime, MiniMax, etc., as well as Supports access to private Al models.