Kimi is an artificial intelligence (AI) chatbot and series of large language models developed by Chinese company Moonshot AI. Its Kimi K3, released July 2026, is the largest open weights model at 2.8 trillion parameters.
Moonshot releases the Kimi series of open weights large language models, available via web, API, and the Kimi Code command-line agent. These are capable of multimodal image and video input, and their "Swarm" feature uses hundreds of sub-agents for additional research.
Despite its open weights, K3's custom license requires revenue sharing of up to 30% for inference providers generating over US$20 million annually.
Kimi K2, a one-trillion parameter model, was released July 2025. It was followed by K2.5 in January 2026, and K2.6 in April. Kimi models have been used as the basis or for post-training of US models, including Thinking Machines Lab's Inkling, Cursor's Composer 2, and others. In September 2026, Anthropic accused Moonshot of covertly routing Kimi requests to its Claude models and distilling Claude via these responses.
Contents
History
Moonshot AI was founded in March 2023 in China. In October 2023, the company officially released the Kimi chatbot and began closed-beta testing.
In July 2024, Kimi's "context caching" feature entered public beta.
On 11 October 2024, Kimi Explore Edition, featuring AI-powered autonomous search, went live globally; monthly active users have since exceeded 36 million.
In September 2025, Moonshot AI added an agentic AI feature to Kimi known as "OK Computer". It is capable of creating multi-page websites and editable slides from simple user prompts and can process up to 1 million rows of input data at once, and output text, audio, images, and video.
In August 2026, legal startup Harvey unveiled its new model using Kimi.
Models
On 16 November 2023, Kimi was released to the general public based on the Moonshot model. The first version of Kimi supported lossless context of 128,000 tokens, making it the first AI model that was capable of accepting contexts of this size.
In March 2024, Moonshot AI began closed beta testing of an updated version of Kimi with a 2 million character context window.
In November 2024, Kimi began internal testing of an AI video generation model.
On 20 January 2025, Kimi K1.5 was released. Moonshot AI claimed it matched the performance of OpenAI o1 in mathematics, coding, and multimodal reasoning capabilities.
In April 2025, the company released Kimi-VL, an open source 16 billion parameter mixture of experts (MoE) large language model with 3 billion active parameters. In June, a reasoning model named Kimi-VL-Thinking was also released.
In June 2025, Kimi-Dev, a 72B parameter coding-focused model based on Qwen2.5-72B, was released. Also in June, Moonshot AI released Kimi-Researcher, an autonomous AI research agent available through Kimi's website and app.
In July 2025, Moonshot AI released Kimi K2, a 1 trillion parameter model with 32 billion active parameters which was open sourced under a modified MIT license. On 9 September 2025, Moonshot AI released an updated version of K2, Kimi-K2-Instruct-0905, which improved its performance in coding tasks and increased its context window from 128K tokens to 256K tokens.
In January 2026, Moonshot AI released Kimi K2.5, a 1 trillion parameter MoE model with 32 billion active parameters. The model is multimodal in vision and language understanding with advanced agentic capabilities, instant and thinking modes, as well as conversational and agentic paradigms. Moonshot AI released Kimi K2.6 in April 2026, followed by Kimi K2.7 Code in June 2026.
Techniques
Kimi Delta Attention (KDA) is an attention mechanism that runs in linear time and constant memory.
In the standard causally masked dot-product attention mechanism, the input sequence of vectors
x
1
,
x
2
,
…
{\displaystyle x_{1},x_{2},\dots }
are converted to a triple sequence of query-key-value vectors
q
i
=
W
Q
x
i
,
k
i
=
W
K
x
i
,
v
i
=
W
V
k