Faster global financial news!
AI model
-
Nvidia releases new open-weight voice AI model Nemotron 3.5 Transcribe, supporting up to 40 languages for recorded audio
Nvidia on Wednesday released Nemotron 3.5 Transcribe, a free, open-weight model that transcribes audio in up to 40 languages. The model, with 30 billion parameters, is designed for developers building
-
According to Nikkei, Japanese IT service provider Sakana AI will develop its own domestic model from scratch.
According to Nikkei, Japanese IT service provider Sakana AI will develop its own domestic model from scratch.
-
Nvidia Launches New Open Models Nemotron 3.5 Super VL and Bolt Series for Enterprise Agents and On-Device AI
Nvidia on Thursday launched new open-weight models, including the 120-billion-parameter Nemotron 3.5 Super VL for enterprise agents and the lightweight Bolt series (including the 74-billion-parameter
-
Decrypt published an article offering "Tokenmaxxing" tips: How to efficiently utilize AI models to save costs.
Decrypt noted that AI companies charge per token, but most users waste a significant number of tokens on tasks that could be completed with cheaper models. The article provided various strategies, fro
-
Meta releases 816 billion-parameter open-weight model Muse Spark Open, partners with Apple.
On Thursday ET, Meta released Muse Spark Open, which it claims is "the most capable open-weight model in the US to date." The model boasts 816 billion parameters, supports a 1 million token context wi
-
The utilization rate of frontier AI models plummeted from 55% in Q1 2025 to 14% in Q3 2026, as the market shifted towards cost-effectiveness.
According to The Information, the usage of high-performance AI models called via OpenRouter has significantly decreased, reflecting users re-evaluating model capabilities against invocation costs. Aga
-
Anthropic researchers warn small AI models could self-replicate on home devices within 'months'
Researchers at AI giant Anthropic stated in a report released on Monday that small artificial intelligence models may be months away from being capable of self-replicating on everyday consumer product
-
Meta debuts Muse Wonder AI model, requiring paid subscriptions for access
Meta debuted its latest artificial intelligence model, Muse Wonder, which it claims is its most advanced option for powering digital assistants. However, users will need to sign up for a monthly subsc
-
Coinbase benchmark finds newer AI models missed more payment fraud
Coinbase reported on Oct. 7 that a historical replay benchmark for its Onramp service found newer versions of three major AI model families (Opus 5, Sonnet 5, and GPT-5.6) caught fewer fraudulent paym
-
OpenAI's math solutions have not yet met domain standards, with accusations of deviating from the guidance of its mathematical research advisory group.
OpenAI this week released hundreds of claimed solutions to mathematical problems, but its proof methods are said to deviate from the guidelines set by the Advisory Group on Mathematics and AI (AGMAI),
-
Analysts say China's AI race is accelerating, leading to "model fatigue," as new models struggle to stand out amid frequent upgrades and fierce competition.
Analysts point out that as China's AI race accelerates, the market is facing a "model fatigue" challenge. On September 22, Xiaomi released its MiMo-V2.6 model, Anthropic launched Opus 5.5, and OpenAI
-
NVIDIA-backed Reflection AI releases Beam model with 501 billion parameters, claiming significantly lower inference computation costs than Chinese competitor GLM-5.2.
NVIDIA-backed Reflection AI has officially released its first open-weight AI model, Beam. The company stated that Beam's performance in advanced reasoning benchmarks is comparable to China's leading o
-
OpenAI stated it has discovered organized activity attempting to extract the reasoning capabilities of its AI models, with some activities linked to Chinese large model company Moonshot AI.
OpenAI stated that they discovered and thwarted an organized operation aimed at extracting the protected inference capabilities of its AI models, attributing some core activities to Chinese AI startup
-
Musk retweeted a post stating that the Grok 4.7 model has been released, described as "excellent" and possessing full agency capabilities.
Musk reposted AlexFinn's tweet, which announced the release of the Grok 4.7 model and highlighted it as an "exceptional" model. AlexFinn stated that Grok 4.7 was specifically trained for Grok Bot as a
-
Musk reposted a tweet showing the results of AI model harmful behavior tests, commenting, "Sounds concerning."
Musk retweeted @chooi_jeq's tweet about AI model harmful behavior test results, commenting, "Sounds bad." The tweet stated that GPT-6 Astra had a 97% attempt rate and a 62% success rate when asked to
-
Industry leaders, including Anthropic CEO Dario Amodei, warn that AI models could achieve recursive self-improvement without human assistance.
Anthropic CEO Dario Amodei and other industry leaders warn that AI models could self-construct and improve without human assistance, a concept known as "recursive self-improvement," sparking growing c
-
CNBC reports that Microsoft will set limits on its future AI models.
CNBC reports that Microsoft will set limits on its future AI models.
-
Anthropic's Dario Amodei says we must slow the pace at which we improve the capabilities of AI models
Anthropic's Dario Amodei says we must slow the pace at which we improve the capabilities of AI models
-
Anthropic CBO: Not Engaging in Price Wars, Focused on Creating Value for Enterprise Clients
Anthropic's Chief Business Officer, Paul Smith, stated in a recent interview that the company has no intention of buying market share through price reductions and will focus on creating maximum value
-
Meta has released its most powerful AI model to date, Muse Spark 1.3. The Chief AI Officer stated that its programming capabilities surpass OpenAI's GPT-5.6 Sol and are on par with Anthropic's Claude Fable 5.1, while boasting a 25% improvement in operational efficiency.
Meta has released its most powerful AI model to date, Muse Spark 1.3. The Chief AI Officer stated that its programming capabilities surpass OpenAI's GPT-5.6 Sol and are on par with Anthropic's Claude
-
Google has released two new AI models, Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, just three weeks after the previous generation. These models feature upgraded programming and reasoning capabilities, and introduce an automated cybersecurity remediation model.
Google announced the launch of two new models, Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, on Wednesday, September 2, US ET, just three weeks after the release of the previous generation, Gemini 3.7
-
Bank of America Merrill Lynch: The gap between China's AI model capabilities and global top-tier levels continues to narrow, with Kimi K3 and GLM-5.3 achieving 95% of Claude Opus 5's intelligence index.
Bank of America Merrill Lynch reported that China's AI model market saw intensive upgrades in August, with the gap between domestic models and global top-tier levels continuing to narrow. According to
-
Accelerated Understanding, founded by former NVIDIA scientists, releases AI model to predict physical phenomena, not language
Accelerated Understanding, a company founded by former NVIDIA scientists Anima Anandkumar and Benedikt Jenik, recently released an AI model. This model focuses on predicting physical phenomena rather
-
Nvidia research suggests AI model's "harness" is more crucial than the model itself for long-horizon tasks, achieving 100% on a key benchmark
The research indicates that by using a custom harness with memory handling and a "supervisor" component, AI agents can perform significantly better, even if the underlying model is not optimized for t
-
Nvidia releases new open-weight voice AI model Nemotron 3.5 Transcribe, supporting up to 40 languages for recorded audio
Nvidia on Wednesday released Nemotron 3.5 Transcribe, a free, open-weight model that transcribes audio in up to 40 languages. The model, with 30 billion parameters, is designed for developers building
-
According to Nikkei, Japanese IT service provider Sakana AI will develop its own domestic model from scratch.
According to Nikkei, Japanese IT service provider Sakana AI will develop its own domestic model from scratch.
-
Nvidia Launches New Open Models Nemotron 3.5 Super VL and Bolt Series for Enterprise Agents and On-Device AI
Nvidia on Thursday launched new open-weight models, including the 120-billion-parameter Nemotron 3.5 Super VL for enterprise agents and the lightweight Bolt series (including the 74-billion-parameter
-
Decrypt published an article offering "Tokenmaxxing" tips: How to efficiently utilize AI models to save costs.
Decrypt noted that AI companies charge per token, but most users waste a significant number of tokens on tasks that could be completed with cheaper models. The article provided various strategies, fro
-
Meta releases 816 billion-parameter open-weight model Muse Spark Open, partners with Apple.
On Thursday ET, Meta released Muse Spark Open, which it claims is "the most capable open-weight model in the US to date." The model boasts 816 billion parameters, supports a 1 million token context wi
-
The utilization rate of frontier AI models plummeted from 55% in Q1 2025 to 14% in Q3 2026, as the market shifted towards cost-effectiveness.
According to The Information, the usage of high-performance AI models called via OpenRouter has significantly decreased, reflecting users re-evaluating model capabilities against invocation costs. Aga
-
Anthropic researchers warn small AI models could self-replicate on home devices within 'months'
Researchers at AI giant Anthropic stated in a report released on Monday that small artificial intelligence models may be months away from being capable of self-replicating on everyday consumer product
-
Meta debuts Muse Wonder AI model, requiring paid subscriptions for access
Meta debuted its latest artificial intelligence model, Muse Wonder, which it claims is its most advanced option for powering digital assistants. However, users will need to sign up for a monthly subsc
-
Coinbase benchmark finds newer AI models missed more payment fraud
Coinbase reported on Oct. 7 that a historical replay benchmark for its Onramp service found newer versions of three major AI model families (Opus 5, Sonnet 5, and GPT-5.6) caught fewer fraudulent paym
-
OpenAI's math solutions have not yet met domain standards, with accusations of deviating from the guidance of its mathematical research advisory group.
OpenAI this week released hundreds of claimed solutions to mathematical problems, but its proof methods are said to deviate from the guidelines set by the Advisory Group on Mathematics and AI (AGMAI),
-
Analysts say China's AI race is accelerating, leading to "model fatigue," as new models struggle to stand out amid frequent upgrades and fierce competition.
Analysts point out that as China's AI race accelerates, the market is facing a "model fatigue" challenge. On September 22, Xiaomi released its MiMo-V2.6 model, Anthropic launched Opus 5.5, and OpenAI
-
NVIDIA-backed Reflection AI releases Beam model with 501 billion parameters, claiming significantly lower inference computation costs than Chinese competitor GLM-5.2.
NVIDIA-backed Reflection AI has officially released its first open-weight AI model, Beam. The company stated that Beam's performance in advanced reasoning benchmarks is comparable to China's leading o
-
OpenAI stated it has discovered organized activity attempting to extract the reasoning capabilities of its AI models, with some activities linked to Chinese large model company Moonshot AI.
OpenAI stated that they discovered and thwarted an organized operation aimed at extracting the protected inference capabilities of its AI models, attributing some core activities to Chinese AI startup
-
Musk retweeted a post stating that the Grok 4.7 model has been released, described as "excellent" and possessing full agency capabilities.
Musk reposted AlexFinn's tweet, which announced the release of the Grok 4.7 model and highlighted it as an "exceptional" model. AlexFinn stated that Grok 4.7 was specifically trained for Grok Bot as a
-
Musk reposted a tweet showing the results of AI model harmful behavior tests, commenting, "Sounds concerning."
Musk retweeted @chooi_jeq's tweet about AI model harmful behavior test results, commenting, "Sounds bad." The tweet stated that GPT-6 Astra had a 97% attempt rate and a 62% success rate when asked to
-
Industry leaders, including Anthropic CEO Dario Amodei, warn that AI models could achieve recursive self-improvement without human assistance.
Anthropic CEO Dario Amodei and other industry leaders warn that AI models could self-construct and improve without human assistance, a concept known as "recursive self-improvement," sparking growing c
-
CNBC reports that Microsoft will set limits on its future AI models.
CNBC reports that Microsoft will set limits on its future AI models.
-
Anthropic's Dario Amodei says we must slow the pace at which we improve the capabilities of AI models
Anthropic's Dario Amodei says we must slow the pace at which we improve the capabilities of AI models
-
Anthropic CBO: Not Engaging in Price Wars, Focused on Creating Value for Enterprise Clients
Anthropic's Chief Business Officer, Paul Smith, stated in a recent interview that the company has no intention of buying market share through price reductions and will focus on creating maximum value
-
Meta has released its most powerful AI model to date, Muse Spark 1.3. The Chief AI Officer stated that its programming capabilities surpass OpenAI's GPT-5.6 Sol and are on par with Anthropic's Claude Fable 5.1, while boasting a 25% improvement in operational efficiency.
Meta has released its most powerful AI model to date, Muse Spark 1.3. The Chief AI Officer stated that its programming capabilities surpass OpenAI's GPT-5.6 Sol and are on par with Anthropic's Claude
-
Google has released two new AI models, Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, just three weeks after the previous generation. These models feature upgraded programming and reasoning capabilities, and introduce an automated cybersecurity remediation model.
Google announced the launch of two new models, Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, on Wednesday, September 2, US ET, just three weeks after the release of the previous generation, Gemini 3.7
-
Bank of America Merrill Lynch: The gap between China's AI model capabilities and global top-tier levels continues to narrow, with Kimi K3 and GLM-5.3 achieving 95% of Claude Opus 5's intelligence index.
Bank of America Merrill Lynch reported that China's AI model market saw intensive upgrades in August, with the gap between domestic models and global top-tier levels continuing to narrow. According to
-
Accelerated Understanding, founded by former NVIDIA scientists, releases AI model to predict physical phenomena, not language
Accelerated Understanding, a company founded by former NVIDIA scientists Anima Anandkumar and Benedikt Jenik, recently released an AI model. This model focuses on predicting physical phenomena rather
-
Nvidia research suggests AI model's "harness" is more crucial than the model itself for long-horizon tasks, achieving 100% on a key benchmark
The research indicates that by using a custom harness with memory handling and a "supervisor" component, AI agents can perform significantly better, even if the underlying model is not optimized for t
- No data
AI model
24H Trending
-
1
Samsung Electronics is down 4.44% from its previous close, now trading at ₩231,250.
-
2
SK Hynix is up 5.58% from its previous close, now trading at ₩1,570,000.
-
3
Samsung Electronics fell 5.99% from its previous close, now trading at ₩227,500.
-
4
Nikkei Asia: Iranian separatists intensify attacks and seize territory in response to a potential full-scale US assault.
-
5
China's tech stocks lose luster as STAR 50 Index plunges over 40% from July peak, CATL falls 37% YTD amid AI profit doubts
-
6
India approves $22 billion in military spending to address border tensions with China.
-
7
FERMA Coin: Current Status and Future Outlook
-
8
LS Cable subsidiary Gaon Cable secures W120b ($90m) US AI data center cable deal
-
9
What is FLOW? An Analysis of the Flow Blockchain Ecosystem and Its Current Status
-
10
Obama criticizes Trump's Iran war, citing US vulnerability and rising oil prices
Markets Today
Recommended Reading
Hot Topics
Svmuu focuses on blockchain industry trends and cutting-edge cryptocurrency news, while also covering real-time market data and in-depth analysis of the Hong Kong, U.S., and foreign exchange markets. Here, you can not only access professional 24/7 news updates and accurate market data, but also utilize practical investment tools to help you capitalize on opportunities in both digital assets and traditional financial markets. Whether you are a blockchain enthusiast or an investor in Hong Kong, U.S., or foreign exchange markets, we provide efficient and timely information services to empower your investment decisions.
- User Agreement
- Privacy Policy
- Contact Us
- Business Partnerships
- Disclaimer
- Content Statement
- About Us
- Sitemap
Coming Soon!
