a16z Top 100: Chất lượng mô hình video AI của Trung Quốc tiếp tục "dẫn đầu", Sora không vào Top 50 di động

The 1M AI News report states that Olivia Moore, a partner at the venture capital firm Andreessen Horowitz (a16z), has released the sixth edition of the Global AI Consumer Product Top 100 list (data as of January 2026), with the AI video generation sector showing the most significant changes. a16z notes that models developed in China are "consistently leading in output quality," with Kling AI, Hailuo, and PixVerse having established actual user bases and made the list. Google's Veo 3 is the first U.S. model to narrow the gap, driving Google Labs' ranking from 36th to 25th.

OpenAI introduced the video generation standalone app Sora 2.0 in September last year, which topped the U.S. App Store for 20 consecutive days and surpassed 1 million downloads faster than ChatGPT's initial release, with current daily active users still exceeding 3 million. However, the download volume has since declined continuously and failed to achieve viral spreading in the form of a social product, resulting in not making it to the Top 50 on mobile devices in this edition.

There has been a significant consolidation in image generation tools. While there were originally seven image generation products in the first edition on the web, now only three remain. Midjourney dropped from the initial Top 10 to 46th place. a16z believes that the remarkable improvement of ChatGPT and Gemini's native image models has sharply raised the bar for standalone products, with Leonardo, Ideogram, and CivitAI serving vertical communities with specific aesthetic preferences. Music and voice are relatively more defensive: the AI music platform Suno maintains its 15th position, while the AI voice platform ElevenLabs has been listed in every edition since its debut in 2023.