{"id":658384,"date":"2026-08-28T11:53:17","date_gmt":"2026-08-28T09:53:17","guid":{"rendered":"https://aivancity.ai/blog/?p=658384"},"modified":"2026-10-07T14:30:34","modified_gmt":"2026-10-07T12:30:34","slug":"kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres","status":"publish","type":"post","link":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/","title":{"rendered":"Kimi K3 : ce qu’il faut savoir du modèle open source de 2 800 milliards de paramètres"},"content":{"rendered":"\n

For a long time, the global race for artificial intelligence seemed to be dominated by American research labs. OpenAI, Anthropic, Google, and xAI held the top spots in international rankings, while Chinese models struggled to compete on the most demanding benchmarks. This dynamic is changing rapidly. Following DeepSeek, GLM-5.2, and Qwen3.7-Max, a new competitor is now grabbing everyone’s attention: Kimi K3, the latest model developed by the Chinese startup Moonshot AI.

\n\n\n\n

With a Mixture of Experts (MoE) architecture comprising 2,800 billion parameters, Kimi K3 aims to rival the most advanced proprietary models on the market while maintaining an open approach. More than just a technical improvement, its launch confirms the spectacular acceleration of China’s artificial intelligence ecosystem and illustrates the gradual narrowing of the gap with U.S. leaders.

\n\n\n\n

Moonshot AI Wants to Shift the Global Balance of Power in AI

\n\n\n\n

Founded in Beijing in 2023, Moonshot AI has quickly established itself as one of China’s most ambitious companies in the field of large language models. After releasing several versions of its Kimi family, the startup is now unveiling Kimi K3, a model designed to meet the needs of both developers and businesses.

\n\n\n\n

Like its main competitors, Kimi K3 can generate code, analyze complex documents, assist with research tasks, automate business processes, and even serve as the foundation for AI agents capable of executing complete workflows.

\n\n\n\n

The goal is clear: to offer a credible alternative to the models developed by OpenAI, Anthropic, or Google, while focusing on making the ecosystem more open.

\n\n\n\n\n\n
\n
\n\n
\n \"MSc\n
\n
\n\n
\n\n \"aivancity\"\n\n
\n

MSc in Generative,
, and Agent-Based AI

\n

Become an expert in generative and agent-based AI: LLMs, transformers, and enterprise deployment. Includes a learning trip to Silicon Valley. RNCP Level 7 certification (equivalent to a 6-year post-secondary degree).

\n\n
\n \n \n \n \n \n 12 months — 6 years of post-secondary education\n \n\n \n \n \n \n \n Master's degree (Bac+5) with experience\n \n\n \n \n \n \n Part-time — Fridays & Saturdays\n \n\n \n \n \n \n \n Paris-Villejuif Campus\n \n
\n
\n\n
\n \n Learn more about the program →\n \n\n \n \n RNCP Level 7 Certification\n \n
\n\n
\n
\n
\n\n\n\n

The Kimi K3 Joins the Ranks of the Top-Performing Models

\n\n\n\n

Just a few days after its launch, Kimi K3 is already among the world's top-ranked artificial intelligence models.

\n\n\n\n

According to the Artificial Analysis Intelligence Index v4.1, which aggregates several leading benchmarks such as Humanity’s Last Exam, GPQA Diamond, SciCode, and Terminal-Bench, Kimi K3 now ranks third globally, just behind Claude Fable 5 and GPT-5.6 Sol. In particular, the model outperforms Claude Opus 4.8, GPT-5.6 Tetra, Grok 4.5, Gemini 3.5 Flash, DeepSeek V4-Pro, and GLM-5.2.1

\n\n\n\n

This improvement is particularly impressive when comparing Kimi K3 to its predecessor. While Kimi K2.6 was still among the mid-tier models, the new version has made a spectacular leap forward and narrowed the gap with the best proprietary AIs on the market to just a few points.

\n\n\n\n
\"\"
Figure 1. Artificial Analysis Intelligence Index v4.1 ranking comparing the leading artificial intelligence models. Kimi K3 ranks third globally, behind Claude Fable 5 and GPT-5.6 Sol. Source: Artificial Analysis / Moonshot AI.
\n\n\n\n

Notable Achievements in Software Development

\n\n\n\n

One reason Kimi K3 attracts so much attention is its performance in software development.

\n\n\n\n

Evaluations published by Moonshot AI show that the model achieves particularly high scores on several specialized code-generation benchmarks. In some tests, it even outperforms several U.S. proprietary models.

\n\n\n\n

Kimi K3 also stands out on Frontend Code Arena, an independent ranking that evaluates models' ability to produce modern, functional web interfaces that developers can use right away. These results reinforce its position as a tool designed for both software engineers and development teams that use coding assistants.

\n\n\n\n

AI Designed for Agents and Automation

\n\n\n\n

Beyond coding, Moonshot AI places a strong emphasis on agent-based applications.

\n\n\n\n

In particular, Kimi K3 achieves one of the highest scores on GDPval-AA v2, a benchmark that measures models’ ability to perform real-world tasks autonomously, ranging from analyzing information to executing complex processes.

\n\n\n\n

This shift confirms a broader trend observed across all major tech companies: AI models are no longer designed solely to answer questions, but are gradually becoming engines capable of powering autonomous assistants, automating business workflows, and interacting with various software tools.

\n\n\n\n

\n

\n
\n
\n \"MSc\n
\n
\n
\n \"aivancity\"\n
\n

MSc in Data Engineering
, and Cloud Computing

\n

Master the entire data and cloud engineering lifecycle (AWS, Azure). RNCP Level 7 certification (equivalent to a 5- or 6-year degree).

\n
\n 2 years — 5 or 6 years of post-secondary education\n Bachelor's degree through Master's degree\n Work-Study Program or Traditional Education\n Paris-Villejuif Campus\n
\n
\n
\n \n Learn more about the program →\n \n \n \n RNCP Level 7 Certification\n \n
\n
\n
\n

\n\n\n\n

An open architecture with 2,800 billion parameters

\n\n\n\n

One of the key features of Kimi K3 is its Mixture of Experts (MoE) architecture.

\n\n\n\n

The model is based on 2,800 billion parameters, but only a subset of these experts is activated for each query. This approach significantly increases the model's capabilities while limiting computational costs during inference.

\n\n\n\n

Moonshot AI also offers Kimi K3 through a so-called \" open\" approach, allowing developers, researchers, and companies to study it, adapt it, or deploy it within their own infrastructure in accordance with the terms defined by the company.

\n\n\n\n

This strategy aligns Kimi K3 with the open-model initiatives spearheaded in recent months by DeepSeek and Qwen and helps make access to very large models more widely available.

\n\n\n\n

China Confirms Its Accelerated Progress in Artificial Intelligence

\n\n\n\n

The launch of the Kimi K3 goes far beyond a simple technology announcement.

\n\n\n\n

In less than two years, several Chinese companies have succeeded in developing models capable of competing with the best international benchmarks. DeepSeek, Alibaba, Zhipu AI, Moonshot AI, and MiniMax are now helping to make China one of the world’s leading centers for research and innovation in artificial intelligence.

\n\n\n\n

This rise in influence is gradually shifting the balance of power in the industry. Whereas U.S. manufacturers once largely dominated international rankings, competition is now much more open, with several Chinese models capable of achieving comparable performance levels on many benchmarks.

\n\n\n\n

A New Milestone in the Global LLM Competition

\n\n\n\n

With Kimi K3, Moonshot AI demonstrates that open models can now come close to matching the performance of the most advanced proprietary systems.

\n\n\n\n

This development is excellent news for developers and businesses, who now have an ever-wider range of high-performance models to choose from based on their technical needs, sovereignty requirements, or deployment costs.

\n\n\n\n

More broadly, Kimi K3 confirms that the next phase of the global competition in artificial intelligence will no longer be limited to OpenAI, Anthropic, and Google. Chinese players are now emerging as competitors capable of having a lasting impact on the evolution of large language models.

\n\n\n\n\n\n
\n\n
\n Technology Framework\n

How does Kimi K3 work?

\n
\n\n

\n Kimi K3 is a large language model (LLM) developed by the Chinese startup Moonshot AI. Designed to compete with the most advanced models from OpenAI, Anthropic, and Google, it is based on a Mixture of Experts (MoE) architecture with 2,800 billion parameters. Unlike traditional dense models, this architecture activates only a portion of the parameters for each query, which improves performance while limiting computational costs. This approach offers a better balance between power, execution speed, and energy efficiency.\n

\n\n

\n Kimi K3 is a versatile model capable of understanding and generating text, writing code, analyzing documents, solving complex problems, and automating tasks. It has been trained on large volumes of multilingual data and optimized for reasoning, programming, and the use of artificial intelligence agents capable of performing a sequence of actions autonomously.\n

\n\n

\n One of the key features of Kimi K3 is its positioning as an open-source model (Open Weights). Moonshot AI makes the model’s weights available so that researchers, developers, and companies can study, adapt, or deploy it in their own environments. This strategy fosters innovation and offers an alternative to the fully proprietary models offered by major U.S. research labs.\n

\n\n

\n Kimi K3’s performance quickly caught the attention of the artificial intelligence community. According to the Artificial Analysis Intelligence Index v4.1, the model ranks among the top three AI systems on the market. It also achieves excellent results on several specialized benchmarks in code generation, web development, and the automation of complex tasks, confirming its focus on professional applications and software development.\n

\n\n
\n
Key Features of the Kimi K3
\n\n \n
\n\n
\n
Technical constraints and limitations
\n\n \n
\n\n
\n\n

\n From a technological standpoint, Kimi K3 confirms the rise of China’s artificial intelligence ecosystem and marks a new milestone in the development of open models. While the first large open models often lagged behind proprietary solutions, Kimi K3 demonstrates that it is now possible to combine openness, top-tier performance, and advanced professional applications.\n

\n\n

\n Its Mixture of Experts architecture also represents a significant advancement in the design of very large models. Rather than utilizing all parameters for every query, the system selects the most relevant experts based on the task at hand. This approach aims to significantly increase the model’s overall capacity while better controlling the resources required for its operation.\n

\n\n

\n Moonshot AI’s “Open Weights” approach ultimately helps diversify the global landscape of large language models. Researchers and companies can more freely experiment with, adapt, and deploy these technologies within their own infrastructure, thereby strengthening technological competition among the world’s leading players in artificial intelligence.\n

\n\n
\n Key Takeaway: Kimi K3 marks a significant milestone in the evolution of open models. Thanks to its Mixture of Experts architecture with 2.8 trillion parameters and its performance in coding, reasoning, and automation, it has established itself as one of the most competitive models on the market. Its launch also illustrates the acceleration of Chinese research in artificial intelligence and the gradual narrowing of the gap with U.S. leaders.\n
\n\n
\n\n
\n\n\n\n

Learn more 

\n\n\n\n

The arrival of the Kimi K3 confirms the rise of open-weight models capable of competing with industry leaders while promoting greater openness in the AI ecosystem. On a related topic, check out our article “ MiniMax’s M3: A New Open-Weight Giant Challenges OpenAI and Anthropic,” which analyzes how a new generation of open models is redefining global competition among AI labs and accelerating innovation in the field of large language models.

\n\n\n\n

References

\n\n\n\n

1. Artificial Analysis. (2026). Artificial Analysis Intelligence Index v4.1.
https://artificialanalysis.ai/

\n","protected":false},"excerpt":{"rendered":"

For a long time, the global race in artificial intelligence seemed to be dominated by American research labs. OpenAI, Anthropic, Google, and xAI held the top spots in international rankings, while Chinese models struggled to compete…

\n","protected":false},"author":7,"featured_media":658567,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"content-type":"","footnotes":""},"categories":[63,43],"tags":[59],"class_list":["post-658384","post","type-post","status-publish","format-standard","has-post-thumbnail","category-ia-agentique","category-innovation-competitivite-par-lia","tag-parlonsia"],"yoast_head":"\nKimi K3: The Chinese AI Giant Challenging OpenAI and Anthropic\n\n\n\n\n\n\n\n\n\n\n\n\n\t\n\t\n\t\n\n\n\n\t\n\t\n\t\n\n","yoast_head_json":{"title":"Kimi K3 : le géant chinois de l’IA qui défie OpenAI et Anthropic","description":"Moonshot AI unveils Kimi K3, an open-source model with 2,800 billion parameters. Ranked third by Artificial Analysis, it rivals Claude Fable 5 and GPT-5.6 Sol in coding, reasoning, and automation.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/","og_locale":"fr_FR","og_type":"article","og_title":"Kimi K3 : le géant chinois de l’IA qui défie OpenAI et Anthropic","og_description":"Moonshot AI dévoile Kimi K3, un modèle open source de 2 800 milliards de paramètres. Classé troisième par Artificial Analysis, il rivalise avec Claude Fable 5 et GPT-5.6 Sol en codage, raisonnement et automatisation.","og_url":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/","og_site_name":"aivancity blog","article_published_time":"2026-08-28T09:53:17+00:00","article_modified_time":"2026-10-07T12:30:34+00:00","og_image":[{"width":1024,"height":1024,"url":"https://aivancity.ai/en/blog/wp-content/uploads/2026/08/lArchitecte-Cloud-2-1.png","type":"image/png"}],"author":"aivancity","twitter_card":"summary_large_image","twitter_misc":{"Écrit par":"aivancity","Durée de lecture estimée":"11 minutes"},"schema":{"@context":"https://schema.org","@graph":[{"@type":"Article","@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#article","isPartOf":{"@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/"},"author":{"name":"aivancity","@id":"https://aivancity.ai/blog/#/schema/person/70f8508e84e45571c5fd172ea40ef3d4"},"headline":"Kimi K3 : ce qu’il faut savoir du modèle open source de 2 800 milliards de paramètres","datePublished":"2026-08-28T09:53:17+00:00","dateModified":"2026-10-07T12:30:34+00:00","mainEntityOfPage":{"@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/"},"wordCount":2282,"image":{"@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#primaryimage"},"thumbnailUrl":"https://aivancity.ai/blog/wp-content/uploads/2026/08/lArchitecte-Cloud-2-1.png","keywords":["Parlons IA"],"articleSection":["IA Agentique","Innovation & compétitivité par l’IA"],"inLanguage":"fr-FR"},{"@type":"WebPage","@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/","url":"https://aivancity.ai/en/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/","name":"Kimi K3 : le géant chinois de l’IA qui défie OpenAI et Anthropic","isPartOf":{"@id":"https://aivancity.ai/blog/#website"},"primaryImageOfPage":{"@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#primaryimage"},"image":{"@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#primaryimage"},"thumbnailUrl":"https://aivancity.ai/blog/wp-content/uploads/2026/08/lArchitecte-Cloud-2-1.png","datePublished":"2026-08-28T09:53:17+00:00","dateModified":"2026-10-07T12:30:34+00:00","author":{"@id":"https://aivancity.ai/blog/#/schema/person/70f8508e84e45571c5fd172ea40ef3d4"},"description":"Moonshot AI unveils Kimi K3, an open-source model with 2,800 billion parameters. Ranked third by Artificial Analysis, it rivals Claude Fable 5 and GPT-5.6 Sol in coding, reasoning, and automation.","breadcrumb":{"@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#breadcrumb"},"inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/"]}]},{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#primaryimage","url":"https://aivancity.ai/en/blog/wp-content/uploads/2026/08/lArchitecte-Cloud-2-1.png","contentUrl":"https://aivancity.ai/blog/wp-content/uploads/2026/08/lArchitecte-Cloud-2-1.png","width":1024,"height":1024},{"@type":"BreadcrumbList","@id":"https://aivancity.ai/blog/kimi-k3-ce-quil-faut-savoir-du-modele-open-source-de-2-800-milliards-de-parametres/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Accueil","item":"https://aivancity.ai/blog/"},{"@type":"ListItem","position":2,"name":"Kimi K3 : ce qu’il faut savoir du modèle open source de 2 800 milliards de paramètres"}]},{"@type":"WebSite","@id":"https://aivancity.ai/blog/#website","url":"https://aivancity.ai/en/blog/","name":"aivancity blog","description":"Advancing Education in Artificial Intelligence","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://aivancity.ai/blog/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https://aivancity.ai/blog/#/schema/person/70f8508e84e45571c5fd172ea40ef3d4","name":"aivancity","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https://secure.gravatar.com/avatar/0b60f844cf48367ece3a9988562f25406b914c56b83ccd3df68e4c07737dc27e?s=96&d=mm&r=g","url":"https://secure.gravatar.com/avatar/0b60f844cf48367ece3a9988562f25406b914c56b83ccd3df68e4c07737dc27e?s=96&d=mm&r=g","contentUrl":"https://secure.gravatar.com/avatar/0b60f844cf48367ece3a9988562f25406b914c56b83ccd3df68e4c07737dc27e?s=96&d=mm&r=g","caption":"aivancity"},"url":"https://aivancity.ai/en/blog/author/bouazizaivancity-ai/"}]}},"amp_enabled":true,"_links":{"self":[{"href":"https://aivancity.ai/blog/wp-json/wp/v2/posts/658384","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https://aivancity.ai/blog/wp-json/wp/v2/posts"}],"about":[{"href":"https://aivancity.ai/blog/wp-json/wp/v2/types/post"}],"author":[{"embeddable":true,"href":"https://aivancity.ai/blog/wp-json/wp/v2/users/7"}],"replies":[{"embeddable":true,"href":"https://aivancity.ai/blog/wp-json/wp/v2/comments?post=658384"}],"version-history":[{"count":4,"href":"https://aivancity.ai/blog/wp-json/wp/v2/posts/658384/revisions"}],"predecessor-version":[{"id":658732,"href":"https://aivancity.ai/blog/wp-json/wp/v2/posts/658384/revisions/658732"}],"wp:featuredmedia":[{"embeddable":true,"href":"https://aivancity.ai/blog/wp-json/wp/v2/media/658567"}],"wp:attachment":[{"href":"https://aivancity.ai/blog/wp-json/wp/v2/media?parent=658384"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https://aivancity.ai/blog/wp-json/wp/v2/categories?post=658384"},{"taxonomy":"post_tag","embeddable":true,"href":"https://aivancity.ai/blog/wp-json/wp/v2/tags?post=658384"}],"curies":[{"name":"wp","href":"https://api.w.org/{rel}","templated":true}]}}