Chen Mian on Evoken, Liblib, Lovat, and LibTV — cash flow, competition, and founder philosophy
Source: LateTalk Episode 175 · Original: LatePost
This is a LatePost Talks interview with Chen Mian, founder of Evoken and the AI products Liblib, Lovat, and LibTV. At a time when fundraising for Chinese AI application companies is extremely difficult, Evoken raised $300 million at a $2 billion valuation, making it one of the highest-valued Chinese AI application companies. The interview is not a funding celebration, but a probe into a harder question: for an application-layer company that does not control the underlying model, how do you define your position, make money, and survive until the endgame while the model-layer narrative is still unfolding?
It is not yet the age of applications; the main narrative is still models and infrastructure. The top priority for an application-layer company is to survive. Source
Chen Mian's judgment: discovering PMF and ultimately owning the market are not equivalent. The real winner must stockpile cash, brand attention, and organizational resilience before the incremental explosion, or else merely paves the way for others.
Entered the AI creator ecosystem as a model community, accumulating early users and model authors.
● ExplorationThe company nearly ran out of cash; technical leaders left one after another. Only a new investment kept it alive.
● Near-deathNew capital came in; the company continued to operate independently instead of being acquired by a giant.
● RescueExtended community capabilities into video generation, searching for a new PMF.
● MigrationLibTV hit a peak daily revenue of $1 million within its first month and became the main revenue contributor.
● GrowthRaised $300 million at a $2 billion valuation; operating cash flow turned positive starting in May.
● Independence