Kunlun AI Video Model Hits #1, Powers $480M Short-Drama Platform

SkyReels V4 Claims Top Spot, Outperforming Google's Veo 3.1

Chinese technology firm Kunlun's SkyReels V4 model has been ranked as the world's leading text-to-video generation model with integrated audio, according to a new report from the independent research firm Artificial Analysis. The model surpassed prominent rivals including Google's Veo 3.1 and Kling 3.0. This ranking is based on blind evaluations from real-world users, providing a more objective measure of performance than internal company benchmarks. The ascent was rapid; after reaching the number two position with a preview version on February 27, 2026, the full V4 model claimed the top spot less than a month later in March.

This marks a significant milestone, positioning a Chinese-developed AI video model at the forefront of a highly competitive global field. SkyReels V4 also demonstrated strong performance in other categories, securing the second-place rank for text-to-video generation without audio. The model's capabilities are set to be officially launched at the 2026 Zhongguancun Forum, with its API already available to developers.

DramaWave Platform Converts Tech Lead into $480M Revenue

Kunlun is successfully translating its technological leadership into commercial results. The SkyReels V4 model is the core engine behind DramaWave, the company's proprietary short-drama streaming platform. As of January 2026, the platform, which also includes FreeReels, reported over 80 million monthly active users (MAU). Financially, it achieved an annualized recurring revenue (ARR) of over $480 million, with monthly revenue hitting $40 million.

The success of DramaWave validates the growing market for AI-assisted short-form video content. This trend is attracting significant investment globally, as evidenced by the recent launch of AdsDrama in the Dominican Republic on March 14, 2026. That new venture, backed by a $50 million investment, aims to build a similar short-drama marketing ecosystem. Kunlun's ability to not only lead in AI development but also capture a large user base and generate substantial revenue places it in a strong competitive position.

Unified Model Solves Key Production Bottlenecks

SkyReels V4's primary advantage lies in its ability to solve critical and long-standing problems in AI video production. Its most significant upgrade is a multi-frame reference capability, allowing creators to input up to nine keyframes to ensure consistent character appearance, clothing, and scenery throughout a video. This directly addresses a major issue that has limited the quality of AI-generated series, where characters often change appearance between scenes.

The model also integrates video generation, editing, and audio synchronization into a single, unified framework. This allows creators to perform complex tasks—such as adding or removing objects, changing character actions, and ensuring accurate lip-syncing—within one system, drastically streamlining the production workflow. For example, one AI-generated short drama produced with Kunlun's tools cost less than $20,000 and went on to generate millions of views, demonstrating the platform's capacity to deliver high-quality content at a fraction of traditional production costs.