Moonshot AI K2 Thinking Model Released and Open-Sourced, with Major Upgrades to Agent and Reasoning Capabilities | BlueRun Ventures Family Headlines
API is now live.
Today, Moonshot AI released Kimi K2 Thinking — the most capable open-source reasoning model Kimi has built to date.
Kimi K2 Thinking is a next-generation Thinking Agent trained by the Kimi team based on the "Model as Agent" philosophy. It natively masters the ability to "think while using tools." It achieves state-of-the-art (SOTA) performance on multiple benchmarks including Humanity's Last Exam, BrowseComp (autonomous web browsing), and SEAL-0 (complex information gathering and reasoning), with comprehensive improvements in agentic search, agentic programming, writing, and general reasoning capabilities.
The Kimi K2 Thinking model can autonomously execute up to 300 rounds of tool calls and sustain stable multi-turn reasoning without human intervention, helping users solve more complex problems. This represents the Kimi team's latest progress in Test-Time Scaling — achieving stronger agent and reasoning performance by simultaneously scaling both reasoning tokens and tool-call iterations.
The Kimi K2 Thinking model is now available on kimi.com and in the regular chat mode of the latest Kimi mobile app. The underlying model for Kimi Agent mode will also be upgraded to Kimi K2 Thinking in the future, bringing full multi-turn reasoning and tool-calling capabilities.
The Kimi K2 Thinking model API is accessible through the Kimi Open Platform (platform.moonshot.cn). For self-hosting, download the model from platforms such as Hugging Face and ModelScope.
BlueRun Ventures is an early investor in Moonshot AI and has continued to increase its stake. We congratulate Moonshot AI on reaching this new milestone. BlueRun also believes more strongly than ever that open-source models like Kimi will drive the industry to spawn more innovative applications and unlock productivity gains through diverse pathways.
The Kimi K2 Thinking model demonstrates powerful reasoning and problem-solving capabilities in Humanity's Last Exam, an ultimate closed-book academic test covering over 100 specialized fields. Under equal conditions with access to tools — search, Python, and web browsing — Kimi K2 Thinking achieved a SOTA score of 44.9% on this benchmark.
Here's an example of the reasoning process for a humanities question from Humanity's Last Exam. In this example, Kimi K2 Thinking conducts 5 rounds of search and reasoning, incorporating new information found in each round to progressively deepen its analysis and ultimately arrive at the answer:
↕ Scroll up and down to view the complete reasoning process
In complex search and browsing scenarios, the Kimi K2 Thinking model also performs exceptionally well. BrowseComp is a benchmark released by OpenAI specifically designed to evaluate AI agent web browsing capabilities. The test was created to measure the persistence and creativity of AI agents in information-overloaded environments — essentially, whether they can "dig deep" like human researchers. On this highly challenging task, human performance averages only 29.2%. Kimi K2 Thinking demonstrated remarkable research tenacity, achieving a new SOTA of 60.2% on this benchmark.
Driven by long-horizon planning and autonomous search capabilities, Kimi K2 Thinking can leverage dynamic loops of up to hundreds of rounds of "think → search → browse webpage → think → code," continuously proposing and refining hypotheses, verifying evidence, conducting reasoning, and constructing logically consistent answers. This ability to actively search while continuously thinking enables Kimi K2 Thinking to decompose vague, open-ended questions into clear, executable subtasks.
Here's an example: through two rounds of search and reasoning, Kimi K2 Thinking first identifies a fastener manufacturing company based on known information about stock buybacks, then locates the stock buyback announcement on the U.S. Securities and Exchange Commission (SEC) website, arriving at an accurate answer:
↕ Scroll up and down to view the complete reasoning process
The Kimi K2 Thinking model's coding capabilities have also been enhanced, with further improvements on benchmarks including SWE-Multilingual (multilingual software engineering), SWE-bench verified set, and Terminal usage.
The Kimi team observed noticeable performance gains for Kimi K2 Thinking when handling HTML, React, and component-rich frontend tasks, transforming creative ideas into fully functional, responsive products. In agentic coding scenarios, Kimi K2 Thinking can think while calling various tools, flexibly integrating into software agents to handle more complex, multi-step development workflows.
Here are two examples:
Now, Kimi K2 Thinking can help you replicate a fully functional Word text editor.
Kimi K2 Thinking can also help you create a flamboyant voxel art piece:
Creative Writing: Kimi K2 Thinking significantly enhances writing capabilities. It can transform rough inspirations into clear, compelling, and purposeful narratives with both rhythm and depth. It easily handles subtle stylistic variations and ambiguous structures while maintaining stylistic coherence across long-form content. In creative writing, its imagery is more vivid, its emotional resonance stronger, blending precise expression with rich performative power.
Academics and Research: In academic research and professional domains, Kimi K2 Thinking shows notable improvements in analytical depth, informational accuracy, and logical structure. It methodically dissects complex instructions and develops ideas with clarity and rigor. This makes it especially adept at handling academic papers, technical abstracts, and lengthy reports demanding high information integrity and reasoning quality.
Personal and Emotional: When responding to personal or emotional questions, Kimi K2 Thinking's answers are more empathetic and balanced in perspective. Its thinking is thorough, thoughtful, and specific, offering nuanced viewpoints and practical follow-up suggestions. It helps users work through complex decisions with clarity and care, its tone grounded, genuinely pertinent, and more human.
Here's an example of assisting with reading an English technical paper:
↕ Scroll up and down to view the complete analysis process
Low-bit quantization is an effective method for reducing latency and GPU memory footprint on large-scale inference servers. The Kimi team's testing found that because reasoning models produce extremely long decoding lengths, conventional quantization methods often lead to significant performance degradation. To overcome this challenge, the Kimi team adopted Quantization-Aware Training (QAT) during the post-training stage and applied INT4 weight-only quantization to MoE components.
This enables the Kimi K2 Thinking model to support native INT4 inference in complex reasoning and agentic tasks, improving generation speed by approximately 2x. INT4 offers stronger compatibility with inference hardware and is more friendly to domestic AI accelerators. Notably, all benchmark results were achieved under INT4 precision.
Go to kimi.com or update to the latest Kimi App, turn on the "Long Thinking" switch for the K2 model from the Toolbox, and throw your complex tasks at Kimi to think through together. The Kimi K2 Thinking model API is now available on the Kimi Open Platform (platform.moonshot.cn), supporting 256K context at the same pricing as Kimi K2-0905: ¥4 per million input tokens, ¥16 per million output tokens, and ¥1 per million cached input tokens. The Turbo API with speeds up to 100 tokens/s is also available simultaneously at ¥8 per million input tokens, ¥58 per million output tokens, and ¥1 per million cached input tokens. Developers are welcome to test and provide feedback on the new model API; please refer to this document for a getting-started guide.
For more model performance evaluation data and use cases, please refer to this technical blog.
$2
About the Kimi K2 Model
The Kimi K2 model was initially released on July 11. It is an open-source foundation model with a Mixture-of-Experts (MoE) architecture, 1 trillion total parameters, and 32 billion active parameters. On September 5, the Kimi K2-0905 update further improved coding capabilities and expanded the context window from 128K to 256K. To date, products including Cline, Cursor, flowith, Genspark, Kilo Code, Kortix Suna, OpenRouter, Perplexity, RooCode, TRAE, Trickle, Vercel, Windsurf, and YouWare have integrated or are using the Kimi K2 model. On November 6, Moonshot AI released the Kimi K2 Thinking model, comprehensively enhancing agent and reasoning capabilities.
Kimi Researcher Team on Agent: It's Not a Frankenstein, It's the Model | BlueRun Family Headlines