Latest news with #GLM-4.5
Time of India
a day ago
- Business
- Time of India
This Chinese company included in US restricted entity list announces ChatGPT's newest rival that claims to be cheaper than DeepSeek
Chinese startup has announced its newest artificial intelligence model, GLM-4.5. The company claims that its latest AI model will be cheaper to use than DeepSeek . This new model shows how Chinese companies are creating more capable AI models at reduced costs. previously known as Zhipu, has confirmed that GLM-4.5 is built on "agentic" AI principles. This means the model can automatically decompose a task into sub-tasks to complete it with greater accuracy, which is a different approach from the logic of some existing AI models. In June, OpenAI included Zhipu in a warning list regarding advancements in Chinese AI. The US government has also placed the startup on its restricted entity list, limiting American firms from engaging in business with it. The new GLM-4.5 model is also open-sourced, which will allow developers to download and use it for free. What said about its AI model GLM-4.5 being cheaper than DeepSeek In a statement to CNBC, Zhang Peng , the CEO of said the company's new GLM-4.5 model runs on eight Nvidia H20 chips . These are Nvidia's AI training chips that are specifically designed for the Chinese market to comply with US export rules. While Nvidia recently received approval to resume shipments to China after a pause, the timeline for delivery remains unclear. Zhang noted that has adequate computing resources and does not need to purchase additional chips for now, but declined to disclose how much was spent on training the model, saying more information would be shared later. said it will price GLM-4.5 at 11 cents per million input tokens and 28 cents per million output tokens, compared to DeepSeek R1's rates of 14 cents and $2.19, respectively. These tokens are used to measure the volume of data processed by AI models. In recent weeks, several Chinese firms have introduced new open-source AI models . At the World AI Conference in Shanghai, Tencent unveiled its HunyuanWorld-1.0 model, designed to help generate 3D scenes for game development. Alibaba followed with the launch of its Qwen3-Coder model, focused on coding tasks. Earlier this month, Moonshot, which Alibaba backs, announced Kimi K2, which it said performs better than OpenAI's ChatGPT and Anthropic's Claude in specific coding tasks. According to the company's website, Kimi K2 charges 15 cents per million input tokens and $2.50 per million output tokens. Back in January, DeepSeek drew attention from global investors with its AI model, which it said was developed despite US chip restrictions and came with lower training and operating costs than its US counterparts. The company claimed its V3 model was trained for under $6 million, though some analysts noted that figure reflects part of its total hardware investment, which exceeded $500 million. iQOO Z10R 5G goes on Sale: BEST Budget Phone for Content Creators?
India Today
a day ago
- Business
- India Today
DeepSeek effect: New free, open source ChatGPT rival GLM 4.5 breaks cover in China
A new AI model has been introduced in China. Called GLM-4.5, this is an open-source model unlike most of the US-based AI systems that are closed, and some benchmarks put it ahead of even DeepSeek R1 and ChatGPT. GLM 4.5 has been developed by startup (formerly known as Zhipu). According to the company, this model is designed specifically for intelligent agent tasks and uses what is described as an 'agentic' AI architecture. This means that the AI model can autonomously take on tasks and handle reasoning, coding and other applications more effectively. advertisementAccording to the company, GLM-4.5 is a large language model featuring 355 billion total parameters, with an optimised variant named GLM-4.5 Air that has 106 billion parameters, making it lighter and faster. The model supports a context window of 128,000 tokens, allowing it to process long conversations or documents without losing focus. It is also said to include native function calling, enabling seamless integration with external software and workflows. The company highlights that these capabilities make GLM-4.5 suitable for a wide range of applications, including advanced coding, physics simulations, game development and interactive company has also revealed that the GLM 4.5 has been released under an Apache 2.0 open-source licence. This means it is free for use and developers can freely download and deploy it, CNBC reported. claims that GLM4.5 ranks third globally and first among Chinese and open-source models across 12 major AI evaluation benchmarks. According to the company, the model scored 98.2 per cent on the MATH500 reasoning test and 91 per cent on the AIME24 challenge. It also delivered 64.2 per cent accuracy on SWE-Bench Verified, a benchmark used for software engineering tasks, and achieved a 90.6 per cent tool-calling success rate, edging out leading by Nvidia chip One of GLM4.5's biggest advantages is cost. CEO ZhangPeng told CNBC that the model can run on just eight Nvidia H20 chips. These chips are designed specifically for the Chinese market under US export controls. This is roughly half the hardware required by DeepSeek's comparable model. ZhangPeng revealed that the company does not currently need to purchase additional chips, indicating that the model already has sufficient computing than DeepSeekThe company reveals that it has also aggressively cut token pricing. will charge $0.11 per million input tokens compared with $0.14 for DeepSeek R1, and $0.28 per million output tokens. This is lower than the $2.19 charged by DeepSeek. Notably, tokens are the standard unit of data measurement for AI context, DeepSeek, the advanced LLM launched earlier this year, is developed by the Chinese startup High Flyer AI and is known for rivalling OpenAI's ChatGPT in natural language understanding and reasoning, while requiring significantly less training cost, reportedly under $6 million. Although GLM-4.5 is about half the size of DeepSeek, it is touted to be using agentic AI design to maintain high accuracy and flexibility in completing tasks with fewer computational the arrival of GLM-4.5 comes at a time when China's AI model development is seeing a rapid surge. By July, Chinese companies had released 1,509 large language models, more than any other country, according to the state-owned Xinhua news agency.- Ends
CNBC
2 days ago
- Business
- CNBC
CNBC Daily Open: Investors are signaling it's time for tariffs to move aside
Stock markets in the U.S. and Europe didn't seem that delighted with the U.S.-European Union trade deal reached over the weekend. The S&P 500 ticked up, but by the barest margin, while the Stoxx Europe 600 fell. Both indexes were trading higher during their respective sessions but had given up those gains as the day ended. For those on the continent, perhaps there was a dawning realization that the agreement wasn't too much in their favor. German Chancellor Friedrich Merz said he would have welcomed further easing of trade, while France's minister for Europe, Benjamin Haddad, said the deal was "unbalanced," according to a Google translation. With U.S. President Donald Trump announcing Monday that he would probably impose a blanket tariff of between 15% and 20% on countries without trade agreements, it's starting to seem like most duties will settle around that level eventually, easing some uncertainty. What's more, economists appear to be revising downward their expectations of the impact tariffs will have on the U.S. economy — so any deals in the future might not trigger rallies, or strong ones at least, on Wall Street. Tariff considerations, then, are on the backburner for now. Investors can turn their attention to Magnificent Seven earnings and data on the U.S. economy coming out the next few days. If all goes well, they might give markets the cheer that was missing on Monday. Global baseline tariff of between 15% and20%. For countries that have not negotiated separate trade agreements with the U.S., Trump said he would likely impose that blanket tariff rate on their exports. But Wall Street doesn't seem as frightened of tariffs anymore. Less than two weeks for Russia to reach a peace deal with Ukraine. That's the new deadline Trump issued to Moscow — if Russia fails to meet it, the U.S. president will implement massive "secondary tariffs" on the country's trade partners, Trump said. India has exported more smartphones to the U.S. than China. In the second quarter, 44% of U.S. smartphone imports were assembled in India, while 25% were from China, according to research firm Canalys. India's and China's share a year earlier were 13% and 61%, respectively. A muted response to the EU deal. On Monday, the S&P 500 closed mostly flat, giving up earlier gains. Asia-Pacific markets fell Tuesday. Shares of Singapore Airlines lost as much as 8% after the carrier reported a 59% slump in profit for its fiscal first quarter. [PRO] Watch this index for signs of a new bull phase. This index, which is calculated differently from the price-weighted S&P 500, gives a better gauge of the health of the entire economy and stock market. China's latest AI model claims to be even cheaper to use than DeepSeek Chinese startup formerly known as Zhipu, announced Monday that its new GLM-4.5 AI model would cost less than DeepSeek to use. At about half the size of DeepSeek's model, GLM-4.5 only needs eight Nvidia H20 chips to operate, CEO Zhang Peng told CNBC on Monday. In late June, OpenAI named Zhipu in a warning about Chinese artificial intelligence progress. The U.S. has also added the startup to its entity list that restricts American companies from doing business with it. —
Business Insider
2 days ago
- Business
- Business Insider
China's Newest AI Model Is Cheaper than DeepSeek
Chinese tech startups are following in DeepSeek's footsteps by releasing new artificial intelligence models that are smarter and cheaper, according to CNBC. One of the most notable is formerly known as Zhipu, which on Monday revealed its GLM-4.5 model. The company says that this model costs less to use than DeepSeek's and works with agentic AI, which is a system that breaks large tasks into smaller steps for more accurate results. Z. ai is also making the model open-source so developers can freely download and use it. Elevate Your Investing Strategy: Take advantage of TipRanks Premium at 50% off! Unlock powerful investing tools, advanced data, and expert analyst insights to help you invest with confidence. What's interesting is that GLM-4.5 is roughly half the size of DeepSeek's competing model and requires only eight Nvidia (NVDA) H20 chips to run. For context, these are the specialized chips made for China to comply with U.S. export controls. CEO Zhang Peng told CNBC that the company already has enough computing power and does not need to buy more chips, although he declined to disclose the model's training costs. Nevertheless, stated that it will charge $0.11 per million input tokens and $0.28 per million output tokens, which is significantly cheaper than DeepSeek's rates of $0.14 and $2.19, respectively. It is worth noting that release adds to a wave of new open-source AI models from China. Indeed, earlier this month, Alibaba-backed Moonshot (BABA) introduced its Kimi K2 model that it claimed was better at coding than OpenAI's ChatGPT and Anthropic's Claude. Tencent (TCEHY) also released its HunyuanWorld-1.0 model for 3D game development, and Alibaba announced Qwen3-Coder for programming tasks. In addition, has raised more than $1.5 billion from investors, which has led U.S. regulators to add the startup to their Entity List that limits American companies from working with it. Which Tech Stock Is the Better Buy? Turning to Wall Street, out of the three stocks mentioned above, analysts think that BABA stock has the most room to run. In fact, BABA's average price target of $151.08 per share implies more than 23% upside potential. On the other hand, analysts expect the least from of $184.91 equates to a gain of 5.6%.
CNBC
2 days ago
- Business
- CNBC
China's latest AI model claims to be even cheaper to use than DeepSeek
BEIJING — Chinese companies are making smarter artificial intelligence models that are increasingly cheaper to use, echoing key aspects of DeepSeek's market-shaking breakthrough. Startup formerly known as Zhipu, announced Monday that its new GLM-4.5 AI model would cost less than DeepSeek to use. In contrast to the logic underlying existing AI models, said its new GLM-4.5 is built on what's known as "agentic" AI, meaning that the model automatically breaks down a task into sub-tasks in order to complete it more accurately. The new model is also open sourced, meaning it is free for developers to download and use. At about half the size of DeepSeek's model, GLM-4.5 only needs eight Nvidia H20 chips to operate, CEO Zhang Peng told CNBC on Monday. That's the chip Nvidia customized for China in order to comply with U.S. export controls. The chipmaker said this month that the U.S. will allow it to resume those China sales after a three-month pause, but it's unclear when those shipments will begin. Zhang said the company doesn't need to buy more of the chips as it has enough computing power for now, but declined to share how much spent on training the AI model. Details will be released later, he said. Back in January, DeepSeek had rattled global investors with its apparent ability to defy U.S. chip restrictions and create an AI model that not only rivaled U.S.-based OpenAI's ChatGPT, but also undercut it in training and operating costs. DeepSeek claimed training costs for its V3 model were less than $6 million, although some analysts said that figure was based on the company's hardware spend of more than $500 million over time. said that for its new GLM-4.5 model, it would charge 11 cents per million input tokens versus 14 for DeepSeek R1; and 28 cents per million output tokens versus $2.19 for DeepSeek. Tokens are a way of measuring data for AI model processing. Earlier this month, Alibaba-backed Moonshot released Kimi K2, which claimed to beat OpenAI's ChatGPT and Anthropic's Claude on certain coding capabilities. Kimi K2 charges 15 cents for every 1 million input tokens, and $2.50 per 1 million output tokens, according to its website. In late June, OpenAI named Zhipu in a warning about Chinese AI progress. The U.S. has also added the startup to its entity list that restricts American companies from doing business with it. launched in 2019 and is reportedly planning an initial public offering in Greater China. The startup has raised more than $1.5 billion from investors including Alibaba, Tencent and Qiming Venture Partners, according to PitchBook. Aramco-backed Prosperity7 Ventures as well as municipal funds from the cities of Hangzhou and Chengdu are also among backers, the database showed. In the last few weeks, several other Chinese companies have announced new, open-source AI models. During the World AI Conference in Shanghai, Tencent released the HunyuanWorld-1.0 model for generating three-dimensional scenes for game development. Last week, Alibaba announced its Qwen3-Coder model for writing computer code.
