Latest news with #Kimi
NBC News
14-07-2025
- Business
- NBC News
Another Chinese AI model is turning heads
BEIJING — The latest Chinese generative artificial intelligence model to take on OpenAI's ChatGPT is offering coding capabilities — at a lower price. Alibaba-backed startup Moonshot released on late Friday night its Kimi K2 model: a low-cost, open source large language model — the two factors that underpinned China-based DeepSeek's industry disruption in January. Open-source technology provides source code access for free, an approach that few U.S. tech giants have taken, other than Meta and Google to some extent. Coincidentally, OpenAI CEO Sam Altman announced early Saturday that there would be an indefinite delay of its first open-source model yet again due to safety concerns. OpenAI did not immediately respond to a CNBC request for comment on Kimi K2. One of Kimi K2′s strengths is in writing computer code for applications, an area in which businesses see potential to reduce or replace staff with generative AI. OpenAI's U.S. rival Anthropic focused on coding with its Claude Opus 4 model released in late May. In its release announcement on social media platforms X and GitHub, Moonshot claimed Kimi K2 surpassed Claude Opus 4 on two benchmarks, and had better overall performance than OpenAI's coding-focused GPT-4.1 model, based on several industry metrics. 'No doubt [Kimi K2 is] a globally competitive model, and it's open sourced,' Wei Sun, principal analyst in artificial intelligence at Counterpoint, said in an email Monday. Cheaper option 'On top of that, it has lower token costs, making it attractive for large-scale or budget-sensitive deployments,' she said. The new K2 model is available via Kimi's app and browser interface for free unlike ChatGPT or Claude, which charge monthly subscriptions for their latest AI models. Kimi is also only charging 15 cents for every 1 million input tokens, and $2.50 per 1 million output tokens, according to its website. Tokens are a way of measuring data for AI model processing. In contrast, Claude Opus 4 charges 100 times more for input — $15 per million tokens — and 30 times more for output — $75 per million tokens. Meanwhile, for every one million tokens, GPT-4.1 charges $2 for input and $8 for output. Moonshot AI said on GitHub that developers can use K2 however they wish, with the only requirement that they display 'Kimi K2' on the user interface if the commercial product or service has more than 100 million monthly active users, or makes the equivalent of $20 million in monthly revenue. Hot AI market Initial reviews of K2 on both English and Chinese social media have largely been positive, although there are some reports of hallucinations, a prevalent issue in generative AI, in which the models make up information. Still, K2 is 'the first model I feel comfortable using in production since Claude 3.5 Sonnet,' Pietro Schirano, founder of startup MagicPath that offers AI tools for design, said in a post on X. Moonshot has open sourced some of its prior AI models. The company's chatbot surged in popularity early last year as China's alternative to ChatGPT, which isn't officially available in the country. But similar chatbots from ByteDance and Tencent have since crowded the market, while tech giant Baidu has revamped its core search engine with AI tools. Kimi's latest AI release comes as investors eye Chinese alternatives to U.S. tech in the global AI competition. Still, despite the excitement about DeepSeek, the privately-held company has yet to announce a major upgrade to its R1 and V3 model. Meanwhile, Manus AI, a Chinese startup that emerged earlier this year as another DeepSeek-type upstart, has relocated its headquarters to Singapore. Over in the U.S., OpenAI also has yet to reveal GPT-5. Work on GPT-5 may be taking up engineering resources, preventing OpenAI from progressing on its open-source model, Counterpoint's Sun said, adding that it's challenging to release a powerful open-source model without undermining the competitive advantage of a proprietary model. Grok 4 competitor Kimi K2 is not the company's only recent release. Moonshot launched a Kimi research model last month and claimed it matched Google's Gemini Deep Research 's 26.9 score and beat OpenAI's version on a benchmark called 'Humanity's Last Exam.' The Kimi research model even got a mention last week during Elon Musk's xAI release of Grok 4 — which scored 25.4 on its own on the 'Humanity's Last Exam' benchmark, but attained a 44.4 score when allowed to use a variety of AI tools and web search. 'Kimi-Researcher represents a paradigm shift in agentic AI,' said Winston Ma, adjunct professor at NYU School of Law. He was referring to AI's capability of simultaneously making several decisions on its own to complete a complex task. 'Instead of merely generating fluent responses, it demonstrates autonomous reasoning at an expert level — the kind of complex cognitive work previously missing from LLMs,' Ma said. He is also author of 'The Digital War: How China's Tech Power Shapes the Future of AI, Blockchain and Cyberspace.'
CNBC
14-07-2025
- Business
- CNBC
Alibaba-backed Moonshot releases new Kimi AI model that beats ChatGPT, Claude in coding — and it costs less
BEIJING — The latest Chinese generative artificial intelligence model to take on OpenAI's ChatGPT is offering coding capabilities — at a lower price. Alibaba-backed startup Moonshot released on late Friday night its Kimi K2 model: a low-cost, open source large language model — the two factors that underpinned China-based DeepSeek's industry disruption in January. Open-source technology provides source code access for free, an approach that few U.S. tech giants have taken, other than Meta and Google to some extent. Coincidentally, OpenAI CEO Sam Altman announced early Saturday that there would be an indefinite delay of its first open-source model yet again due to safety concerns. OpenAI did not immediately respond to a CNBC request for comment on Kimi K2. One of Kimi K2's strengths is in writing computer code for applications, an area in which businesses see potential to reduce or replace staff with generative AI. OpenAI's U.S. rival Anthropic focused on coding with its Claude Opus 4 model released in late May. In its release announcement on social media platforms X and GitHub, Moonshot claimed Kimi K2 surpassed Claude Opus 4 on two benchmarks, and had better overall performance than OpenAI's coding-focused GPT-4.1 model, based on several industry metrics. "No doubt [Kimi K2 is] a globally competitive model, and it's open sourced," Wei Sun, principal analyst in artificial intelligence at Counterpoint, said in an email Monday. "On top of that, it has lower token costs, making it attractive for large-scale or budget-sensitive deployments," she said. The new K2 model is available via Kimi's app and browser interface for free unlike ChatGPT or Claude, which charge monthly subscriptions for their latest AI models. Kimi is also only charging 15 cents for every 1 million input tokens, and $2.50 per 1 million output tokens, according to its website. Tokens are a way of measuring data for AI model processing. In contrast, Claude Opus 4 charges 100 times more for input — $15 per million tokens — and 30 times more for output — $75 per million tokens. Meanwhile, for every one million tokens, GPT-4.1 charges $2 for input and $8 for output. Moonshot AI said on GitHub that developers can use K2 however they wish, with the only requirement that they display "Kimi K2" on the user interface if the commercial product or service has more than 100 million monthly active users, or makes the equivalent of $20 million in monthly revenue. Initial reviews of K2 on both English and Chinese social media have largely been positive, although there are some reports of hallucinations, a prevalent issue in generative AI, in which the models make up information. Still, K2 is "the first model I feel comfortable using in production since Claude 3.5 Sonnet," Pietro Schirano, founder of startup MagicPath that offers AI tools for design, said in a post on X. Moonshot has open sourced some of its prior AI models. The company's chatbot surged in popularity early last year as China's alternative to ChatGPT, which isn't officially available in the country. But similar chatbots from ByteDance and Tencent have since crowded the market, while tech giant Baidu has revamped its core search engine with AI tools. Kimi's latest AI release comes as investors eye Chinese alternatives to U.S. tech in the global AI competition. Still, despite the excitement about DeepSeek, the privately-held company has yet to announce a major upgrade to its R1 and V3 model. Meanwhile, Manus AI, a Chinese startup that emerged earlier this year as another DeepSeek-type upstart, has relocated its headquarters to Singapore. Over in the U.S., OpenAI also has yet to reveal GPT-5. Work on GPT-5 may be taking up engineering resources, preventing OpenAI from progressing on its open-source model, Counterpoint's Sun said, adding that it's challenging to release a powerful open-source model without undermining the competitive advantage of a proprietary model. Kimi K2 is not the company's only recent release. Moonshot launched a Kimi research model last month and claimed it matched Google's Gemini Deep Research 's 26.9 score and beat OpenAI's version on a benchmark called "Humanity's Last Exam." The Kimi research model even got a mention last week during Elon Musk's xAI release of Grok 4 — which scored 25.4 on its own on the "Humanity's Last Exam" benchmark, but attained a 44.4 score when allowed to use a variety of AI tools and web search. "Kimi-Researcher represents a paradigm shift in agentic AI," said Winston Ma, adjunct professor at NYU School of Law. He was referring to AI's capability of simultaneously making several decisions on its own to complete a complex task. "Instead of merely generating fluent responses, it demonstrates autonomous reasoning at an expert level — the kind of complex cognitive work previously missing from LLMs," Ma said. He is also author of "The Digital War: How China's Tech Power Shapes the Future of AI, Blockchain and Cyberspace."
Yahoo
04-07-2025
- Automotive
- Yahoo
Wimbledon daily highlights
Why we had a tricky Spielberg weekend | 2025 Austrian GP F1 Akkodis Race Debrief A challenging weekend for the team in Spielberg. Andrew Shovlin, our Trackside Engineering Director, is here to answer your questions from the Austrian GP in our Akkodis Race Debrief. From whether Kimi is okay, to why we struggled in the Austrian heat and much more. 7:17 Now Playing Paused Ad Playing
eNCA
16-06-2025
- Automotive
- eNCA
Antonelli comes of age with podium finish in Canada
MONTREAL - Italian teenager Kimi Antonelli raised the biggest and most emotional reaction on Sunday when he came of age as a Formula One star by finishing third with a mature race for Mercedes at the Canadian Grand Prix. The 18-year-old rookie, who passed his road driving test earlier this year before the season began, became the third-youngest podium finisher in Formula One history. Four-time world champion Max Verstappen, who was second for Red Bull in Sunday's race behind Mercedes team leader George Russell, was aged 18 years, seven months and 15 days old when he took his first podium in 2016. Lance Stroll was also slightly younger than Antonelli when he came third in Azerbaijan in 2017. "It was so stressful for me," he said. "But I am super happy!" Beaming from ear to ear, Antonelli waved to the big crowd at the Circuit Gilles Villeneuve amid chants of "Kimi, Kimi" as they applauded his appearance on the podium after the race. "I had a good start. I managed to jump into P3 and I just stayed up there at the front. The last stint, I pushed a bit too hard behind Max and I killed a bit of the front left tyre and then I struggled a bit at the end. "So, I'm really happy to bring the podium home for the team." He started the race in thrilling style by passing championship leader Oscar Piastri's McLaren for third place on the opening lap. "I had a good first launch and managed to get alongside and then, in the first corner, I just tried to carry as much speed as possible. When I was alongside him in Turn Two, I knew I would have the advantage going into Turn Three." He said he hoped his team's first win this year and since the Las Vegas race last year would be the springboard for more victories. But Russell was quick to play down too many expectations, pointing out that the Canada track was one that was very suitable to the strengths of their Mercedes car. "It's a little bit cooler around here," said Russell. "We saw it last year in Canada. I would love to get our hopes up, but I think the strength of our car is in the cooler conditions so let's see in the coming races. "But, yeah, we will enjoy this for now." "Time will tell. We won Austria last year so maybe that's a good omen. We did have high expectations coming into this weekend and it worked out as we thought."
News18
10-06-2025
- News18
Lights Out For AI: China's Chatbots Go Dark To Keep College Entrance Exam Fair
Last Updated: The Gaokao, running from June 7 to 10 this year, is given by over 13.4 million students who are vying for limited university spots in China. China's artificial intelligence chatbots have temporarily hit the off switch on some key features, turning a little less smart for a very serious reason: preventing cheating during the massive Gaokao university entrance exams. Leading tech giants like Alibaba (with its Qwen bot), Tencent (Yuanbao), ByteDance (Doubao), and Moonshot (Kimi) are among those that have pulled the plug on their most popular AI tools' photo-recognition capabilities as authorities are making sure no student gets an unfair advantage during the high-stakes national tests. The Gaokao, running from June 7 to 10 this year, is given by over 13.4 million students who are vying for limited university spots in China. China has earlier also gone all out with anti-cheating measures, from banning electronic devices to flying surveillance drones. ByteDance's Doubao chatbot, a rival in the AI arena, reportedly still permits image uploads but refuses to answer any test-related queries, citing 'non-compliance with rules." Meanwhile, Alibaba's Qwen bot also reportedly refrains from analyzing test papers during exam hours. This temporary suspension aligns with a broader regulatory framework for generative AI in education that the Chinese Ministry of Education previously released. The guidelines specifically prohibit students from independently utilizing artificial intelligence tools that generate open-ended content within primary and secondary school systems. Instead, educators are encouraged to integrate AI as a supplementary tool to studies, rather than a replacement for human-led instruction. First Published: June 10, 2025, 21:33 IST
