- Moonshot AI’s Kimi K3 targets US$2B in annualised revenue.
- Kimi K3 mixes open weights with commercial monetisation.
Chinese AI company Moonshot AI is targeting US$2 billion in annualised revenue by the end of 2026, putting new figures behind the commercial growth of its open-weight Kimi models.
The company told investors that annual recurring revenue exceeded US$1 billion in August, up from US$300 million in June, Bloomberg reported, citing people familiar with the matter. The US$2 billion figure represents a year-end annualised revenue target rather than revenue booked across the full year.
Bloomberg previously reported that Moonshot was generating more than US$200 million in annual recurring revenue in April, citing financial adviser HF Capital. The revenue came from subscriptions to its Kimi chatbot and AI model services.
The figures show that Moonshot is generating recurring revenue from Kimi, although the financial information reported so far does not disclose its operating profit, gross margins, or the cost of serving the workloads behind that revenue.
Moonshot released Kimi K3 in July. The model uses a mixture-of-experts architecture with 2.8 trillion total parameters, of which 104 billion are activated during inference, and supports a context window of just over one million tokens, according to Moonshot’s model documentation.
Moonshot describes K3 as an open-weight model. Its model weights can be downloaded and deployed outside Moonshot’s services, subject to conditions in the company’s Kimi K3 licence.
How Moonshot monetises K3
Moonshot’s K3 licence allows users to distribute, modify, deploy, and create derivative works from the model, but places commercial conditions on some large-scale deployments.
A company operating a Model-as-a-Service business must enter into a separate agreement with Moonshot before using K3 commercially if the company and its affiliates generate more than US$20 million in aggregate revenue over any consecutive 12-month period. The licence defines Model-as-a-Service as providing third parties with access to model inference or fine-tuning through services such as an API.
The provision does not apply to internal enterprise deployments or products that only embed K3 capabilities within a broader service. Large commercial products using K3 are also subject to attribution requirements if they exceed 100 million monthly active users or US$20 million in monthly revenue.
Reuters reported in August that Moonshot was in early-stage negotiations with Microsoft, Amazon, and Google to host K3 on their cloud platforms, with the company seeking as much as a 30% share of revenue generated from K3-related services. The talks also covered data access and mechanisms for auditing token usage.
The same report said Moonshot had already reached similar revenue-sharing arrangements with smaller cloud providers, including Chinasoft International.
The discussions remained active into September, when Moonshot was still negotiating with Microsoft, Amazon, and Google as the company considered its public listing plans.
Distribution extends beyond Moonshot
As of September 14, OpenRouter listed 16 providers serving Kimi K3, including DigitalOcean and DeepInfra. The model is also available through Moonshot’s own services.
OpenRouter allows different companies to host the same model and routes requests between those providers. Activity through independently hosted versions of K3 therefore does not necessarily represent API revenue paid directly to Moonshot.
As of the same date, OpenRouter listed K3 at about US$2.40 per million input tokens and US$12 per million output tokens, although pricing varies by provider.
Moonshot is selling K3 access alongside Chinese competitors offering lower-priced APIs. DeepSeek’s V4.1 Flash costs US$0.15 per million uncached input tokens and US$0.60 per million output tokens during off-peak periods, with peak rates of US$0.30 and US$1.20 respectively.
Alibaba Cloud lists Qwen3.5 Plus from its Singapore region at US$0.40 per million input tokens and US$2.40 per million output tokens for requests of up to 256,000 input tokens. Prices increase for longer context lengths.
The companies do not publicly disclose enough information on inference costs, customer discounts, or margins to compare the profitability of these services.
Revenue growth meets infrastructure costs
Moonshot also faced computing capacity constraints shortly after K3’s launch. Reuters reported in July that the company temporarily stopped accepting new Kimi subscriptions after requests approached the limits of its existing computing clusters.
Moonshot allocated available capacity to existing paid users and said future membership plans would include a separate coding-focused tier. The pause came shortly after the launch of K3.
Moonshot has also continued raising external capital. Bloomberg reported that the company raised about US$2 billion in May at a valuation above US$20 billion, followed by a US$3.5 billion round in July that valued Moonshot at US$35 billion. The July round exceeded its initial target of between US$1 billion and US$2 billion.
By September, Moonshot had raised more than US$5.5 billion since its founding and had confidentially filed for a Hong Kong initial public offering that could seek about US$3 billion, Reuters reported, citing people familiar with the filing.
Goldman Sachs, China International Capital Corp, and Deutsche Bank are working with Moonshot on the proposed IPO. Reuters also reported that the company was valued at about US$50 billion in an ongoing funding round at the time of the filing.
The available financial information does not show how much it costs Moonshot to generate its reported revenue. Without figures for inference costs, gross margins, operating expenses, or net income, its annual recurring revenue does not establish whether the company is profitable.
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
Tech Wire Asia is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Moonshot’s Kimi growth tests the business case for open-weight AI appeared first on TechWire Asia.