Moonshot AI

Moonshot was founded in March 2023 by Yang Zhilin, Zhou Xinyu and Wu Yuxin who were schoolmates at Tsinghua University. It was launched on the 50th anniversary of Pink Floyd's The Dark Side of the Moon which was Yang's favorite album and the inspiration for the company's name. Yang has stated his goal for founding Moonshot AI is to build foundation models to achieve AGI. Yang's three milestones are long context length, multimodal world model, and a scalable general architecture capable of continuous self-improvement without human input. In January 2026, Moonshot released Kimi K2.5, a multimodal upgrade to Kimi K2 that added native vision capabilities through a 400-million-parameter vision encoder called MoonViT. The model can process both images and video, enabling agentic tasks such as replicating website user journeys from video demonstrations alone. The K2 model was released just 3 months prior. In February 2026, Anthropic accused Moonshot of violating the terms of service by using thousands of fraudulent accounts to generate millions of conversations with Claude to train its own large language models. == Funding and investments ==

Moonshot was valued at $300 million when it received its initial funding of $60 million and had 40 employees. In February 2024, Alibaba Group led a $1 billion funding round for Moonshot, which gave it a valuation of $2.5 billion.Alibaba acquired 36% stake and invested US $0.8 billion in Moonshot AI. Moonshot in January 2026 upgraded its flagship model and now had capabilities that outperformed three of the top U.S. AI models. Alibaba also revealed its latest generative AI model, Qwen3 Max Thinking, which also claimed to have outdone its U.S rivals on a broad benchmark test called 'Humanity's Last Exam. In August 2024, Tencent and Gaorong Capital joined as investors in a $300 million funding round that valued Moonshot at $3.3 billion. In October 2025, Moonshot was reportedly nearing the completion of a new funding round of approximately $600 million, led by IDG Capital with participation from existing investors including Tencent, valuing the company at $3.8 billion pre-money. In March 2026, Moonshot was considering an initial public offering on the Hong Kong Stock Exchange. == Products and research ==

Kimi In October 2023, Moonshot launched its first AI chatbot, Kimi, whose name comes from Yang's English nickname. It had emerged as the closest rival to Baidu's Ernie Bot. In March 2024, Moonshot claimed Kimi could handle 2 million Chinese characters in a single prompt which was a significant upgrade from the previous version that could only handle 200,000. Due to the increased number of users, on 21 March, Kimi suffered an outage for two days and Moonshot had to issue an apology. As of August 2024, Kimi ranked third in active monthly users according to aicpb.com. On 20 January 2025, Kimi K1.5 was released. Moonshot claimed it matched the performance of OpenAI o1 in mathematics, coding, and multimodal reasoning capabilities. In June 2025, Kimi dropped in popularity to seventh place in active monthly users. The model uses a mixture-of-experts (MoE) architecture, where 32 billion parameters are active during inference. K2 was trained on 15.5 trillion tokens of data and is released under a modified MIT license. Kimi K2 is an open source LLM, meaning that it can be downloaded and built upon by users. The day after its release, Kimi K2 had the most downloads on the platform, an increase in popularity from previous months. The release of Kimi K2 follows a trend amongst Chinese companies to make their AI models open sourced likely trying to counter US's efforts to limit China's tech growth. In China, Kimi has six tiers of plans ranging from 5.2 yuan for four days to 399 yuan for a year of priority use. Mooncake serving platform Mooncake is the platform that serves Moonshot's Kimi chatbot and processes 100 billion tokens daily. Moonshot was awarded the Erik Riedel Best Paper Award at the USENIX FAST conference for the paper detailing the architecture of Mooncake. The researchers indicate that Muon improves computational efficiency by a factor of 2 compared to the standard optimizer, AdamW, in training large models. The researchers note that long context scaling and improved policy optimization methods were key, without relying on complex techniques like Monte Carlo tree search, value functions, and process reward models. == See also ==