But the spirit can’t become Seedance? The commercialization divide of China’s AI video duo
36Kr published an article on July 6 discussing the differentiated paths of Byte Seedance and Kuaishou Keling: Seedance relies on the Volcano Engine to monetize B-side MaaS, while Keling relies on the Kuaishou ecosystem to focus on C-side creators. The two companies have different business models and endowments, and they do not have to compete in the same dimension.
On July 6, 2026, 36Kr published an analysis article "Ke Ling Can't Become Seedance". The subtext of the title is actually a rhetorical question: Ke Ling doesn't have to become Seedance. The core of the discussion in the article is the two completely different paths taken by Byte and Kuaishou Keling (Kling) on the AI video track.
Two roads: B-side MaaS and C-side creator economy
Seedance's path is very clear: relying on the Volcano Engine to monetize the B-side of "Model as a Service (MaaS)", its monthly revenue has exceeded 1 billion yuan. The API is priced at "1 yuan per second" and continues to increase the revenue target - this is a path of "selling models, selling computing power, and selling clouds". Keling relies on the Kuaishou ecosystem and focuses more on C-end creators and content ecology. At the same time, it is reported that it plans to spin off and promote an IPO - this is a path of "building products, building communities, and building traffic".
There is no difference between the two routes, but they are inevitable choices determined by endowment. Byte has the cloud base of Volcano Engine, which is naturally suitable for MaaS; Kuaishou has the creator ecology and content distribution, which is naturally suitable for C-side tools and communities.
The world's first echelon under the "Three Kingdoms Killing" pattern
According to 36Kr's "AI Generated Video "Three Kingdoms Killing"", China's AI video track currently presents a pattern of "Seedance taking the lead, Keling spinning off and planning an IPO, and Alibaba following suit". Overseas, players such as OpenAI Sora, Google Veo 3, and Runway are competing.
It is worth emphasizing that Chinese manufacturers are already in the first echelon in the world in terms of AI video generation capabilities and commercialization pace. Seedance has a monthly revenue of 1 billion and Keling is preparing for an IPO. These actions have almost no benchmark among overseas peers - most overseas AI video companies are still in the stage of financing and burning money, while China's top players have entered the stage of "self-development" and "capital marketization".
From an industry perspective, the "division" between Seedance and Keling has an enlightenment for latecomers: the competition in AI video is no longer a competition of pure model capabilities, but a competition of business models and ecological endowments**. Choosing a path and maximizing your talents is more important than catching up on someone else's track. For investors, the two paths correspond to the two valuation logics of "cloud and model commercialization" and "content and creator platform" respectively, and are worthy of separate evaluation.
Several directions worth tracking in the future:
- The progress of Keling’s spin-off and IPO: The listing will give the “AI video C-end company” a public market price for the first time.
- Variables for Alibaba’s entry into the game: What kind of competitive landscape will the Tongyi video model disrupt?
- Achievement of Byte MaaS goals: Can the revenue target of 15 billion yuan be accelerated with the launch of overseas APIs?
- Interpenetration between C-side and B-side: Whether Keling will supplement the B-side API, and whether Seedance will make C-side products.
Reviews