Which company has the #3 AI model end of July? (Style Control On)

The launch of Kimi K3, an artificial intelligence model developed by the Chinese startup Moonshot AI, reignited the debate about China's ability to overcome its limited access to advanced chips and match the performance of US models. Operating similarly to GPT and Claude, this system represented a breakthrough with the potential to transform a sector previously dominated by the United States.

Google Releases Three New A.I. Models

Kimi K3 Release: Is This the DeepSeek 2.0 Moment for US Stocks and Crypto Markets?. • Introduction Kimi K3, Moonshot AI’s 2.8-trillion-parameter open model released on July 16-17, 2026, has already sparked immediat

Following 'Kimi', China's Alibaba Also Unveils New AI... "Ranked 2nd in the World After Claude"

Chinese AI company Moonshot AI surprised markets with the launch of its new Kimi K3 model, reigniting the debate about the US's progress in the AI ​​race compared to China. The company stated that the new model outperforms most global competitors, with the exception of advanced models from Anthropic and OpenAI, suggesting that China may be closing the gap with the US at a faster pace than anticipated, according to Bloomberg. Previous estimates indicated that Chinese AI models lagged behind leading American models by about 6 to 9 months, but some experts believe the gap has now narrowed to just two or three months in certain areas, particularly programming. Ion Stoica, a computer science professor at the University of California, Berkeley, and co-founder of Databricks, said the recent development reflects a significant shift in the ability of Chinese open-source models to compete with advanced models. The launch of Kimi K3 has raised concerns in Silicon Valley about the effectiveness of the massive spending by US tech companies on developing AI models, amidst Questions arise about the ability of US companies to maintain their dominance. This new development comes at a time when companies like OpenAI and Anthropic are facing pressure to reduce the costs of developing and using artificial intelligence, with some customers looking for less expensive alternatives. The Kimi K3 model is based on a massive architecture with 2.8 trillion parameters, and Moonshot says it is designed to perform advanced tasks, particularly in programming and proxy AI applications. Analysts believe this new Chinese development could increase global competition in the AI ​​market and put additional pressure on US companies to prove their superiority in terms of performance, cost, and user experience. #Arabic_Business

Chinese company Moonshot AI has launched its open-source artificial intelligence model, Kimi K3, asserting that it rivals the latest models from American companies at a lower cost, with capabilities for data analysis, software development, website creation, video editing, and efficient execution of complex tasks.

Containing 2.8 trillion pieces of information, Chinese company Moonshot announces its latest AI model, the Kimi K3.

China's Moonshot Unveils New AI 'Kimi K3'... Closely Chasing OpenAI and Antropic

A year ago, the big three was OpenAI, Anthropic, and Google. Things have changed. • • Moonshot's Kimi K3 sits above Gemini on every composite benchmark, and it's open source in 10 days. • • New episode: what K3 reveals about frontier margins, model sizes, and who's actually still in the game. • • 00:11 Is Kimi K3 the Third Best Model? • 04:04 Why Delay the Weights? • 05:30 2.8T Parameters and Serving Constraints • 06:48 Frontier Margins and the 3x Price Hike • 11:10 New Architecture, What Comes Next • 14:09 Will Open Source Catch Closed? • 19:51 Built for Chinese Accelerators • 22:57 The Harness Is the Product • 28:49 We're Still Early

A year ago, the big three was OpenAI, Anthropic, and Google. Things have changed. • • Moonshot's Kimi K3 sits above Gemini on every composite benchmark, and it's open source in 10 days. • • New episode: what K3 reveals about frontier margins, model sizes, and who's actually still in the game. • • 00:11 Is Kimi K3 the Third Best Model? • 04:04 Why Delay the Weights? • 05:30 2.8T Parameters and Serving Constraints • 06:48 Frontier Margins and the 3x Price Hike • 11:10 New Architecture, What Comes Next • 14:09 Will Open Source Catch Closed? • 19:51 Built for Chinese Accelerators • 22:57 The Harness Is the Product • 28:49 We're Still Early

Chinese AI company Moonshot launches new model 'Kimi K3'... "Chasing Claude and GPT"

Chinese AI model Kimi K3 has caused shares of global industry leaders to plummet. Shares of companies in the semiconductor and AI sectors fell significantly after the presentation of a new product by Chinese startup Moonshot, Bloomberg reports. Moonshot said that the Kimi K3 model is comparable in capabilities to leading developments from OpenAI and Anthropic. The company also claims that its AI outperforms some competing models in a number of tests. Investors expressed concerns that the rapid development of Chinese AI developments could reduce the need for large-scale investments by American companies in computing infrastructure and advanced chips. Max | Telegram | Newsletter | iOS App | Android

China's Moonshot AI claims Kimi K3 can rival OpenAI and Anthropic

Chinese AI Startup Moonshot Unveils Kimi K3 Model—Will It Challenge OpenAI And Anthropic?. • The Chinese startup called its AI model the “new frontier of intelligence.”

A NEW AI MODEL FROM CHINA JUST GOT WALL STREET WORRIED • • Moonshot's new Kimi K3 model claims to rival the strongest offerings from OpenAI and Anthropic, and investors are already calling it a new "Kimi moment" similar to last years "DeepSeek moment" per Bloomberg. • • The fear isn't really about Kimi K3 itself. It's what one portfolio manager called the capex domino effect: if US firms lean more on Chinese models and less on Western labs, spending on the most advanced chips could slow. • • Moonshot priced K3 at roughly the same level as Anthropic's Sonnet, a premium versus other Chinese models, betting on capability over pure cost. • • Artificial Analysis reportedly ranked K3 ahead of Anthropic's Opus 4.8 on some frontier benchmarks, the first Chinese open-weight model to hit that mark.

From “News Items” from @EllisItems this AM ( • • “2. Financial Times: • • Chinese AI start-up Moonshot has released a large language model with capabilities approaching those of frontier US labs such as Anthropic, as the gap narrows between the two countries on state-of-the-art AI. • The Beijing-based group released Kimi K3, China’s largest AI model to date with 2.8 trillion parameters, on Thursday. The number of parameters refers to the size of the model’s neural network, with a higher count generally leading to greater capabilities. • • Anthropic has not disclosed its models’ parameters, but industry experts speculate its flagship Claude Opus 4.8 has 1.5 trillion-2 trillion parameters. • People with knowledge of the development said K3 would be freely available to download as a so-called open-weight model. • • K3 was beating Opus 4.8 and OpenAI’s GPT 5.5 in most coding and general AI agent benchmarks, according to results released by Moonshot. On most benchmarks it was still falling short of Fable, a powerful model that Anthropic briefly suspended after the US raised concerns over its hacking capabilities. (Source:

From “News Items” from @EllisItems this AM ( • • “2. Financial Times: • • Chinese AI start-up Moonshot has released a large language model with capabilities approaching those of frontier US labs such as Anthropic, as the gap narrows between the two countries on state-of-the-art AI. • The Beijing-based group released Kimi K3, China’s largest AI model to date with 2.8 trillion parameters, on Thursday. The number of parameters refers to the size of the model’s neural network, with a higher count generally leading to greater capabilities. • • Anthropic has not disclosed its models’ parameters, but industry experts speculate its flagship Claude Opus 4.8 has 1.5 trillion-2 trillion parameters. • People with knowledge of the development said K3 would be freely available to download as a so-called open-weight model. • • K3 was beating Opus 4.8 and OpenAI’s GPT 5.5 in most coding and general AI agent benchmarks, according to results released by Moonshot. On most benchmarks it was still falling short of Fable, a powerful model that Anthropic briefly suspended after the US raised concerns over its hacking capabilities. (Source:

Chinese startup Moonshot AI has unveiled a new model it says closes the gap with leading U.S. offerings and surpasses OpenAI and Anthropic’s most capable systems on some benchmarks. • • Kimi K3 still trails Anthropic’s Claude Fable 5 and OpenAI’s GPT 5.6 Sol on overall performance, the company said on Friday, but consistently outperformed other tested models. • • Read more here:

Moonshot is Scheduled to Release Kimi K3, China’s Largest AI Model to Date with 2tn-3tn Parameters, in the Coming Days-FT

$META: Muse Spark 1.1 • • From testing it today, I am very bullish on Muse Spark 1.1. It's nothing like the Llama Models and I would say comparable to OpenAI/Anthropic for a decent sized subset of my usage. Overall, the model accuracy is frontier worthy and the RL on it is almost magical as it's very intuitive akin to Fable 5. • • One surprising thing is that Muse Spark is blazing fast, almost so fast that I wonder if more people adopt it would we see it slow down. But really I think Meta has an vertical integration advantage as it's essentially it's own cloud. This makes me wonder if the best AI Models will all become their own Clouds and be co-designed with their own ASICs which we are seeing from OpenAI and Anthropic already. • • I think Muse Spark can serious become a top 3 model edging out Gemini as Meta has the best RL team and data which is critical for models nowadays. Having Facebook, Instagram, WhatsApp and Messenger user data is an unfair advantage. • • Stock Research • I re-queried the stock research tasks that I had asked ChatGPT-5.5 Pro yesterday on Muse Spark instead. The results were: • • ChatGPT-5.5 Pro • 1. Runs for longer • 2. More detail oriented • • Muse Spark Thinking • 1. Very Intuitive - get's me the information I need and understands concepts with less prompts. It's like asking for RCA and actually getting RCA. • 2. Lightning fast - I'm on my company's enterprise subscription for ChatGPT and Free Muse Spark Thinking got my task done 5x faster. • • Coding • For coding, most developers are used to having a harness so I would like to see a "Muse Code". It's not natively supported by Cursor so Meta should take a lead here and come up with a harness. Coding is where some of the most complex AI usage is and would be useful for Meta to collect data and feedback. • • Synergy With Core Business • Although I am not a Facebook or Instagram user, I tested it out by querying for videos and extracting information and was very impressed. I had expected Grok to be superiorly integrated to X but it felt like agentic search like ChatGPT or Claude querying the web. • • Muse Spark, on the other hand, is much more tightly integrated with Facebook and Instagram and able to perceive more deeply into the content. • • Shopping • Yes, I do use AI to research for buying items like baby products. Previously using ChatGPT but Muse Spark does a way better job because it's way more intuitive. I wonder if Meta has some secret sauce from it's ad targeting as the three modes are "instant", "thinking" and "shopping". This would be pretty big as one of the previous strong holds for Google is that many people still use it over ChatGPT/Claude for shopping/shopping research. However I got to say Muse Spark wins here and it's a very lucrative AI segment.