Core Summary
The recent financing valuation of “Moon’s Dark Side” has soared to $50 billion before investment, but the valuation logic in the AI industry has fundamentally changed: from competing on who has the best model performance to competing on who can turn that model into real cash and provide unique value. Its main challenge now is that it can no longer rely solely on the label of being an “excellent AI assistant” and must break away from the shadow of the “Chinese version of ChatGPT” to establish its own independent narrative and prove its irreplaceability in the AI industry; otherwise, the current high valuation may not be sustainable.
Detailed Analysis
1. Valuation Has Rose Rapidly, but Market Expectations for AI Companies Have Changed
The valuation of Moon’s Dark Side has increased tenfold since the beginning of the year, and it is now seeking a Pre-IPO round with a target pre-investment value of $50 billion. However, the primary and secondary markets are showing a stark contrast:
- Primary Market (Unlisted): OpenAI maintains a valuation of $1 trillion, Anthropic is accelerating its IPO process, and DeepSeek’s valuation continues to rise.
- Secondary Market (Listed): The stock price of Zhipu has fallen by 65% from its peak, while MiniMax has dropped by 84%.
This indicates that the criteria for capital market judgment have changed. In the past, a leading model performance was enough to secure a high valuation, but now everyone’s model capabilities are becoming increasingly similar. Simply having strong technology is no longer sufficient; the market is asking: How much actual revenue can your model generate? Do you have a stable user base and business model?
2. Model Capability Is the “Entry Ticket,” but Long-Term Success Depends on Profitability
After ChatGPT’s emergence in 2022, model performance (such as multimodality and reasoning abilities) was at the core of valuation. However, models are evolving so quickly that what is leading today may be surpassed by others tomorrow, making it increasingly difficult to establish a competitive barrier solely through technology.
- Example of Anthropic: It has generated significant revenue through its enterprise API business (charged based on usage), with annual recurring revenue (ARR) exceeding $60 billion. Its valuation of $1.2 trillion corresponds to an ARR multiple of only 20 times (valuation ÷ ARR).
- Pressure for Moon’s Dark Side: Its ARR is around $300 million, so a $50 billion valuation would correspond to an ARR multiple of 166 times—more than eight times that of Anthropic. This means the market has high expectations for it, but if its revenue growth does not keep up, the valuation could become inflated.
In conclusion, model capability is just the entry requirement; commercialization ability determines how long you can stay in the game.
3. To Achieve a High Valuation, You Need Your Own “Unique Story”
In the capital market, an “independent narrative” (a unique value proposition) determines the upper limit of valuation, while trying to mimic others will only result in a basic valuation. For example:
- OpenAI focuses on the story of “super applications leading to AGI (Artificial General Intelligence), positioning itself as the platform for future intelligence.
- Anthropic emphasizes the reliability of its AI solutions for enterprises.
- DeepSeek highlights its low-cost, open-source models, aiming to reduce training costs.
Moon’s Dark Side once became widely used as an AI assistant thanks to Kimi’s long-text processing capabilities, but this label is now outdated (many products can handle long texts). It needs a new narrative that differentiates it from both the “Chinese version of ChatGPT” and Anthropic by identifying its unique value.
4. Moon’s Dark Side’s New Moves: Moving from C to B Clients? Open Source May Be a Signal
Moon’s Dark Side recently open-sourced the Kimi K3 model. The founder previously stated that closed-source development was the only way to create a super app, so this move may indicate a shift towards B2B (business-to-business) services:
- C-side Challenges: AI services are costly and depend on usage volume; more users and frequent use lead to higher costs, creating economies of scale issues.
- B-side Advantages: Enterprise customers are willing to pay, providing stable revenue sources (e.g., 75% of Anthropic’s revenue comes from enterprise APIs).
However, open-sourcing is just the first step. Moon’s Dark Side also needs to clarify what unique advantages it offers to attract businesses, such as “higher-performance models at lower costs” or solutions that better meet the needs of Chinese companies.
5. The Pressure of High Valuation: Revenue Growth Must Keep Up
Moon’s Dark Side’s ARR multiple of 166 times is higher than that of Anthropic (20 times), Zhipu (62 times), and MiniMax (24 times). This suggests that the market expects it to generate substantial revenue in the future. If its revenue growth slows, the valuation could decline. It needs to demonstrate to the market what additional value it offers, such as new business models, unique technologies, or methods for rapid revenue expansion (e.g., a developer ecosystem).
Final Conclusion
Moon’s Dark Side currently has a high valuation, but it faces significant challenges. It must break away from its previous identity as an “AI assistant” and establish its own independent narrative to convert its model capabilities into stable revenue. Only by doing so can it sustain its current high valuation and gain a solid foothold in the AI competition.