Moonshot AI Raked In More Revenue in 20 Days Than All of Last Year! How to Price AI Products When You're Winning Global Paying Users | Rong Talk --- **[Note: The user provided a headline/fragment for translation. Here is the translated version following all specified rules.]**

On January 27, Moonshot AI released and open-sourced its K2.5 model — the company's most intelligent and capable model to date, with strong visual understanding, coding, and multi-agent capabilities.

The powerful model directly drove explosive revenue growth for Moonshot AI.

On February 23, less than a month after the K2.5 launch, Moonshot AI's cumulative revenue over the past 20 days had already exceeded its total revenue for all of 2025, driven by surging global paid users and API call volume. According to OpenRouter rankings, K2.5 has maintained a leading position in call volume, and ranked first on OpenClaw's model usage leaderboard.

Moonshot AI has once again proven that competitive models and products don't just bring user growth — they directly unlock global revenue. Moonshot AI's revenue saw a "structural reversal" after the K2.5 launch, with overseas revenue now exceeding domestic revenue.

As AI products expand into more global markets, challenges are becoming increasingly prominent: pricing and settlement, cost and risk control, globalization and compliance — all are determining how far a product can go.

Recently, several experts from Stripe, the global financial payments infrastructure platform, shared with Gaorong Ventures portfolio companies how AI products can develop future-proof monetization and growth strategies. Stripe provides financial payments infrastructure for some of the world's fastest-growing AI companies, including OpenAI, Cursor, Midjourney, Manus, and Lovable, giving it a broad global perspective and extensive case studies and data, along with genuinely practical advice.

💡 Speakers

Ariel Lim — Director of Enterprise Accounts, Greater China, Stripe

Lina Shen — Solutions Architect, Financial Automation, Asia-Pacific, Stripe

Haiyue Nie — Business Development Manager, Greater China, Stripe

Here are Stripe's 10 recommendations for AI product commercialization and pricing strategy.

1. Commercialization moves must keep pace with AI technology development.

The number of companies breaking through from $1 million to $10 million ARR is growing rapidly; AI-driven revenue growth is faster than any previous technology wave. Previously, "SaaS 100" companies took an average of 37 months to grow from $500,000 to $5 million in revenue; "AI 100" companies needed 24 months; while the latest "new AI" companies require only 9 months.

This shows that users' willingness to pay for AI products that genuinely solve their problems is much stronger than we imagined, requiring AI product commercialization to keep pace with AI development.

2. AI products are born global — be ready for global customers from day one.

Once an AI product goes live today, it's already facing global users. The top 100 AI companies on Stripe's platform cover twice as many countries in their first year of operation as SaaS companies do. In the past, there was more focus on European and American markets, but many developing economies — such as India, Latin America, and Southeast Asia — also have large numbers of heavy users extremely enthusiastic about AI products and applications.

At the same time, as the AI growth race intensifies, refining core product features and value alone is no longer sufficient to retain users. Peripheral experiences — such as payments and subscription management — are becoming new decisive factors. If you want to earn global money, don't let payments become a stumbling block to growth. AI products' trial-to-paid conversion rates are nearly double those of traditional software, meaning that if the payment process isn't smooth, or if a payment fails at a critical moment, the loss isn't just one order — you're directly handing hard-won users to competitors.

Additionally, regulatory and tax compliance across different regional markets, anti-fraud efforts, and other tasks are becoming increasingly complex.

3. The "growth flywheel" for AI company expansion.

Train models to optimize products >> Reach more global users, get more and faster feedback to update products, while generating more profit >> Raise more capital for infrastructure such as compute power >> Tap global talent pools >> Expand technological boundaries, explore the future of AGI

4. Five important strategic levers for AI product global growth and monetization.

1) Build exceptional PLG

Optimize every touchpoint to reduce friction, improve conversion rates, and increase revenue.

2) Accelerate globalization

Don't negotiate payments country by country; instantly access all major global payment methods.

3) Expand revenue streams

Don't stick to just one pricing model; flexible models open entirely new revenue pathways.

4) Prevent abuse and fraud

AI can help you write code, and it can also help hackers commit fraud. Using AI-driven tools to prevent risks is the only way to protect profits. Keep in mind that for every $1 lost to fraud, companies bear $3.03 in operational costs.

5) Obtain unified customer views

Use a central command center to understand what customers are actually thinking.

5. How should AI products actually be priced? Follow the five-step framework.

When AI companies begin their commercialization journey, they can follow these five steps to build an initial monetization and pricing strategy suitable for most markets, while maintaining flexibility. The core of this framework is that when users pay, they do so genuinely and willingly; when companies collect revenue, they do so with confidence. Ultimately, evolve from selling tokens or traffic to selling value.

1) Value metrics

Clarify what value the product delivers to users — what exactly are users willing to pay for. For example, workflow automation, capability enhancement, or direct revenue improvement.

2) Billing metrics

Determine which billing unit best correlates with product value — consumption-based, workflow-based, or outcome-based. Of course, outcome-based is the most advanced approach.

3) Pricing model

Includes fixed-price subscriptions, pure usage-based, or hybrid pricing models.

4) Billing logic

The specific billing unit used to charge customers — fixed fees, credit allowances, pay-as-you-go, and pay-for-outcomes.

5) Safeguards

This is something many founders might overlook. Implement safety measures to prevent massive bills, such as usage caps, automatic notifications, and rate limiting. This protects not just users, but also the company.

6. Why is "hybrid pricing" the inevitable trend? It balances predictability and expansion potential.

Unlike SaaS products where marginal costs are nearly zero, AI products have increasing marginal costs to serve new users. If AI products adopt pure token-based pricing, it may create usage anxiety for users; but with fixed monthly fees, high infrastructure costs (compute, GPUs, model calls), heavy users, and even freeloaders can erode profit margins.

Hybrid pricing emerges as the natural solution. On one hand, fixed fees provide predictable recurring revenue and stable cash flow; on the other hand, dynamic components are layered on top — for example, overage charges for heavy users. Heavy users consume more because the product meets their needs, and they pay willingly. This is the most scientifically balanced point for both supply and demand.

Data also shows that high-growth AI companies are more inclined toward hybrid pricing. 57% of ultra-high-growth global AI companies adopt hybrid pricing, significantly higher than the 26% adoption rate among low-growth AI companies.

Faced with potentially diverse pricing models, Stripe provides AI companies with a one-stop solution, enabling businesses to activate multiple monetization models and billing logics like building blocks, with no-code or low-code implementation.

7. A tool framework: testing whether your AI product monetization and pricing strategy is sound.

This framework is called the GAPS framework.

1) Growth

Based on current pricing strategy, is it sufficiently easy for users to increase usage or adopt additional products and features?

2) Adoption

Is the adoption barrier low enough and smooth enough? Can users experience the product's value in the shortest time and most effective way?

3) Predictability

Does it bring predictable monthly revenue for the business?

4) Safeguards

Does the pricing and billing model have the ability to resist malicious behavior and potential risks? Can the profit protection net hold up?

8. Pricing models are never set in stone — don't fear changes and adjustments.

Traditional SaaS companies typically consider pricing model changes only every 18 months; but for AI companies, pricing changes are recommended every six months to a year. This is because markets are changing, usage calculations are changing, technology costs are changing, and user perceptions are changing.

The fastest-growing AI companies are constantly adjusting their pricing models, and user acceptance of pricing updates is already quite high. Companies can and should communicate changes to users clearly and transparently.

Excellent commercial teams must possess the ability to experiment quickly, get rapid feedback, and adjust quickly. Don't fear mistakes — if the current pricing model isn't working, quickly adjust and optimize based on data and user feedback.

9. A common mistake AI products make when going global: pricing "not bold enough."

Overseas users have very mature payment psychology. When pushing AI products overseas, why not open a pay button on Day 1? You can absolutely do free trial plus pro subscription from the start, rather than limiting yourself from the beginning.

If your product is good enough, users aren't that sensitive to price. Companies can absolutely provide users with multiple subscription tier options, creating more possibilities for market validation and revenue generation.

10. Agentic Commerce is becoming reality.

As agents evolve, more and more products are embedding AI-native transaction capabilities — allowing agents to pay and place orders themselves. "Agentic Commerce" is gradually becoming an irreversible trend.

Stripe has already partnered with ChatGPT, Microsoft Copilot, and others to support users making payments and transactions directly within conversations. Behind this, Stripe has developed the Agent Commerce Protocol (ACP) and Agent Commerce Suite (ACS), continuously完善ing the economic infrastructure for the agent era.

AI technology and products are still iterating rapidly, and correspondingly, AI product monetization pathways and pricing strategies must maintain the broadest global perspective and greatest flexibility, allowing products to operate freely in the globalization process and convert technological leadership into real revenue growth.

🦞 Bonus

Ronghui Preview: Claude Code and OpenClaw Closed-Door Session

During the Spring Festival, Moonshot AI launched Kimi Claw, enabling one-click deployment. Users don't need to purchase hardware or servers, nor input code, to quickly use OpenClaw-related functions.

On March 7, Gaorong Ventures' Ronghui will hold a closed-door session in Beijing, with Moonshot AI, Pine, Amazon Web Services, and SiliconFlow jointly providing in-depth analysis of the transformative opportunities brought by Claude Code and OpenClaw, and sharing the latest product practices.

👇 AI founders are welcome to scan the QR code to register (scroll down for the full poster; registration requires screening, and selected participants will receive directed meeting instructions)