Kimi-maker Moonshot AI targets $2 billion in annual revenue
TechCrunch Russell Brandom ● Covered by 2 sources
Moonshot AI says its Kimi model could hit $2 billion in yearly revenue by year-end. That’s still tiny next to OpenAI and Anthropic, and its open weights make the math harder.
Based on reporting by TechCrunch, Russell Brandom — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Moonshot AI thinks its open-weight Kimi business can grow fast enough to reach $2 billion in annualized revenue by the end of the year, Bloomberg reported on Friday. That would be twice the company’s reported revenue run rate for August. For a lab best known for a model that people can actually look under the hood of, that is a very ambitious target.
The bet rests on K3, which Moonshot released this summer. Usage has slipped a bit in recent months, but OpenRouter data still shows as many as 300 billion tokens being generated each day by K3 models on the platform. That kind of volume helps explain why Moonshot is talking like a company with a serious sales engine, not just a research shop with a popular demo.
And yet the numbers still sit far below the giants. Recent reports put OpenAI at $40 billion and Anthropic at $65 billion, both well ahead of Moonshot’s target. The difference is structural as much as commercial: Moonshot’s model weights are freely available, so its margins are far thinner than those of closed-weight rivals.
That leaves Moonshot in an awkward but instructive spot. The company is showing that open-weight AI can still make real money. At the same time, its growth story is shadowed by a very different headline: Anthropic recently accused Moonshot of a long-running distillation effort that allegedly sent nearly 300,000 Kimi requests to Claude Opus and gathered more than 23 million responses from Anthropic models for training. The business case for open models may be real. The rules around how some of them get built are still a mess.
My take — AI-written commentary, not fact-checked reporting
Open-weight AI keeps getting sold as the clean, noble alternative, then the incentives arrive and everyone starts freelancing with someone else’s model. Moonshot’s revenue push proves the market is there; the Anthropic accusation shows why the trust gap is, too. Free weights are great until the bill comes due in somebody else’s lab notebook.
Read more about this at: TechCrunch
