7*24

Hot topic

Market

Activities

On June 3, according to Onchain Lens monitoring, a wallet associated with Anchorage Digital staked 55,594 ETH to the Ethereum Beacon Chain, valued at approximately $109.9 million based on current prices.

[Anchorage Digital-Linked Wallet Stakes 55,594 ETH to Ethereum Beacon Chain, Valued at Approximately $109.9 Million]On June 3, according to Onchain Lens monitoring, a wallet associated with Anchorage Digital staked 55,594 ETH to the Ethereum Beacon Chain, valued at approximately $109.9 million based on current prices.

Anchorage Digital-Linked Wallet Stakes 55,594 ETH to Ethereum Beacon Chain, Valued at Approximately $109.9 Million

On June 3, AI startup Special completed a new funding round led by the well-known venture capital firm Andreessen Horowitz (a16z). Sources indicate that this financing will provide the company with the necessary funds to further expand product development and business growth. Special focuses on the field of artificial intelligence, and the specific amount of financing, company valuation, and other participating investors have not yet been disclosed. This investment also indicates that top venture capital firms are actively positioning themselves in emerging AI startups as the generative AI trend continues to drive interest.

[AI Startup Special Secures New Funding Round Led by a16z, Amount and Valuation Undisclosed]On June 3, AI startup Special completed a new funding round led by the well-known venture capital firm Andreessen Horowitz (a16z). Sources indicate that this financing will provide the company with the necessary funds to further expand product development and business growth. Special focuses on the field of artificial intelligence, and the specific amount of financing, company valuation, and other participating investors have not yet been disclosed. This investment also indicates that top venture capital firms are actively positioning themselves in emerging AI startups as the generative AI trend continues to drive interest.

AI Startup Special Secures New Funding Round Led by a16z, Amount and Valuation Undisclosed

On June 3, international oil prices rose on Wednesday due to the escalating situation between the US and Iran and the uncertain prospects for negotiations. Specifically, WTI crude oil futures for July delivery increased by over 1% to $94.81 per barrel, while Brent crude oil futures for August delivery rose by 0.88% to $96.84 per barrel. The US Central Command reported that US forces intercepted multiple Iranian missiles and drones and conducted defensive strikes against Iranian targets. Meanwhile, former President Trump and US Secretary of State Rubio stated that negotiations between the US and Iran are still ongoing, with Iran potentially open to discussions regarding aspects of its nuclear program. However, Iranian media previously reported that there had been no communication between the two sides for several days and suggested that Tehran might seek to fully close the Strait of Hormuz. Analysts from Fitch Group noted that the US-Iran conflict has caused widespread damage to the oil and gas industry in the Middle East, with export disruptions, capacity shutdowns, and repeated attacks on infrastructure resulting in billions of dollars in losses, particularly impacting Qatar, Bahrain, and Iraq the most severely.

[International Oil Prices Rise Over 1% Amid Escalating US-Iran Conflict and Increased Negotiation Uncertainty]On June 3, international oil prices rose on Wednesday due to the escalating situation between the US and Iran and the uncertain prospects for negotiations. Specifically, WTI crude oil futures for July delivery increased by over 1% to $94.81 per barrel, while Brent crude oil futures for August delivery rose by 0.88% to $96.84 per barrel. The US Central Command reported that US forces intercepted multiple Iranian missiles and drones and conducted defensive strikes against Iranian targets. Meanwhile, former President Trump and US Secretary of State Rubio stated that negotiations between the US and Iran are still ongoing, with Iran potentially open to discussions regarding aspects of its nuclear program. However, Iranian media previously reported that there had been no communication between the two sides for several days and suggested that Tehran might seek to fully close the Strait of Hormuz. Analysts from Fitch Group noted that the US-Iran conflict has caused widespread damage to the oil and gas industry in the Middle East, with export disruptions, capacity shutdowns, and repeated attacks on infrastructure resulting in billions of dollars in losses, particularly impacting Qatar, Bahrain, and Iraq the most severely.

International Oil Prices Rise Over 1% Amid Escalating US-Iran Conflict and Increased Negotiation Uncertainty

On June 3, Sensor Tower estimated that ChatGPT, developed by OpenAI, has surpassed 1 billion monthly active users globally, making it the fastest application to reach this milestone. According to Sensor Tower, ChatGPT achieved 1 billion monthly active users in May, approximately three years after its launch, surpassing the previous growth records of Google Maps, TikTok, Instagram, and YouTube.

[ChatGPT Surpasses 1 Billion Monthly Active Users Worldwide]On June 3, Sensor Tower estimated that ChatGPT, developed by OpenAI, has surpassed 1 billion monthly active users globally, making it the fastest application to reach this milestone. According to Sensor Tower, ChatGPT achieved 1 billion monthly active users in May, approximately three years after its launch, surpassing the previous growth records of Google Maps, TikTok, Instagram, and YouTube.

ChatGPT Surpasses 1 Billion Monthly Active Users Worldwide

On June 3, on-chain analyst Yu Jin reported that as ETH dropped to around $1,870, the largest ETH long position on Hyperliquid, which was established at an average price of $2,261 with 120,000 ETH (valued at approximately $271 million), has now seen its unrealized loss expand to about $47 million. Data shows that the liquidation price range for this position is between $1,617 and $1,355. The most recent liquidation price is only about $250 away from the current price, and continued market declines may further increase its liquidation risk.

[Hyperliquid's Largest ETH Long Position Faces $47 Million Unrealized Loss]On June 3, on-chain analyst Yu Jin reported that as ETH dropped to around $1,870, the largest ETH long position on Hyperliquid, which was established at an average price of $2,261 with 120,000 ETH (valued at approximately $271 million), has now seen its unrealized loss expand to about $47 million. Data shows that the liquidation price range for this position is between $1,617 and $1,355. The most recent liquidation price is only about $250 away from the current price, and continued market declines may further increase its liquidation risk.

Hyperliquid's Largest ETH Long Position Faces $47 Million Unrealized Loss

On June 3, according to HTX market data, Bitcoin has risen above $67,000, with the 24-hour decline narrowing to 5.84%.

[Bitcoin Rises Above $67,000]On June 3, according to HTX market data, Bitcoin has risen above $67,000, with the 24-hour decline narrowing to 5.84%.

Bitcoin Rises Above $67,000

On June 3, the Nikkei 225 index surpassed 68,000 points for the first time, with a daily gain expanding to nearly 2%. (Jin Shi)

[Nikkei 225 Index Surpasses 68,000 Points for the First Time, Daily Gain Expands to Nearly 2%]On June 3, the Nikkei 225 index surpassed 68,000 points for the first time, with a daily gain expanding to nearly 2%. (Jin Shi)

Nikkei 225 Index Surpasses 68,000 Points for the First Time, Daily Gain Expands to Nearly 2%

On June 3, the US Central Command announced that US forces shot down multiple Iranian ballistic missiles and drones, and conducted a defensive strike on a military control station on Iran's Qeshm Island. The US stated that the missiles fired by Iran towards Kuwait and Bahrain did not hit their targets, with some intercepted by US and local air defense systems. The Iranian Revolutionary Guard claimed that it used missiles and drones to attack the US Fifth Fleet headquarters, regional air bases, and helicopter facilities in response to previous US attacks on communication facilities in southern Qeshm Island and Iranian tankers near the Strait of Hormuz. As the conflict continues into its third month, neither side has managed to transform the fragile ceasefire into a long-term peace agreement. Iranian media reported that Tehran is evaluating a ceasefire proposal put forth by the Trump administration, but recent communication between the two sides has been limited; Trump stated that US-Iran negotiations are still ongoing. Meanwhile, Kuwait and Bahrain have issued security alerts in response to the escalating missile and drone threats.

[US-Iran Military Actions Escalate, Ceasefire Agreement Under Pressure, Peace Talks Stalled]On June 3, the US Central Command announced that US forces shot down multiple Iranian ballistic missiles and drones, and conducted a defensive strike on a military control station on Iran's Qeshm Island. The US stated that the missiles fired by Iran towards Kuwait and Bahrain did not hit their targets, with some intercepted by US and local air defense systems. The Iranian Revolutionary Guard claimed that it used missiles and drones to attack the US Fifth Fleet headquarters, regional air bases, and helicopter facilities in response to previous US attacks on communication facilities in southern Qeshm Island and Iranian tankers near the Strait of Hormuz. As the conflict continues into its third month, neither side has managed to transform the fragile ceasefire into a long-term peace agreement. Iranian media reported that Tehran is evaluating a ceasefire proposal put forth by the Trump administration, but recent communication between the two sides has been limited; Trump stated that US-Iran negotiations are still ongoing. Meanwhile, Kuwait and Bahrain have issued security alerts in response to the escalating missile and drone threats.

US-Iran Military Actions Escalate, Ceasefire Agreement Under Pressure, Peace Talks Stalled

According to monitoring by Beating, Mustafa Suleyman, head of Microsoft AI (MAI), announced the launch of a new self-developed family of MAI native models at the Build 2026 developer conference. This series includes seven models covering inference, programming, image processing, transcription, and speech, all trained from scratch by Microsoft without using any third-party models for knowledge distillation, and all datasets have been obtained with compliant authorization. Microsoft stated its commitment to building 'Humanist Superintelligence' to ensure that cutting-edge AI serves as a tool for humans under human supervision. Additionally, Microsoft's GB200 computing cluster has been fully deployed to drive the continuous iteration of this model ecosystem. The flagship inference model of the MAI family, MAI-Thinking-1, features 35 billion active parameters, employs a mixture of experts (MoE) architecture, and provides a 128K context window. In mainstream software engineering and mathematical reasoning assessments such as SWE-bench Pro, this model achieved performance comparable to Claude Opus 4.6 and outperformed Claude Sonnet 4.6 in blind human evaluations. For programming scenarios, MAI introduced the MAI-Code-1-Flash, an agent programming model with 5 billion parameters, which will be deeply integrated into GitHub Copilot and VS Code, offering performance on par with Claude Haiku at a lower inference cost. In the multimodal domain, MAI-Image-2.5 and its Flash variant support high-precision text-to-image generation and image editing, with image quality scores surpassing Nano Banana Pro. For speech and transcription, MAI launched the MAI-Transcribe-1.5, a state-of-the-art transcription model for 43 languages with five times the speed of competitors, as well as the MAI-Voice-2 and its Flash variant, which supports 15 languages and features emotion control and zero-shot cloning. These models will not only be deployed on Azure AI Foundry but will also be available on OpenRouter, Fireworks, and Baseten, with developers being able to fine-tune weights for the first time. Microsoft also revealed a 1.4 times improvement in computational efficiency achieved through joint software-hardware optimization with its self-developed chip Maia 200. In addition to the basic model releases, Microsoft launched the 'Frontier Tuning' service based on a Reinforcement Learning Environment (RLE). This service allows enterprises to customize train MAI models using their internal operational trajectories, decision sequences, and specialized data in a fully controlled isolated environment ('training gym'). Tests indicate that custom models after Frontier Tuning significantly improve efficiency, with the MAI model optimized for Excel aligning with GPT-5.4 while being ten times more efficient. Furthermore, a custom MAI model for McKinsey achieved the highest win rate while reducing costs by nearly ten times. Additionally, Microsoft announced a strategic partnership with the Mayo Clinic, a leading global medical institution, to jointly develop a clinical reasoning large model based on Mayo's clinical data and Microsoft's AI foundation. This model will be owned by the Mayo Clinic and will first be deployed internally for early diagnosis and treatment planning, before being made available to other medical institutions through Azure AI Foundry.

[Microsoft AI Launches First Inference Model MAI-Thinking-1 and Six New Native Models, Introduces Private Reinforcement Learning 'Frontier Tuning' Service]According to monitoring by Beating, Mustafa Suleyman, head of Microsoft AI (MAI), announced the launch of a new self-developed family of MAI native models at the Build 2026 developer conference. This series includes seven models covering inference, programming, image processing, transcription, and speech, all trained from scratch by Microsoft without using any third-party models for knowledge distillation, and all datasets have been obtained with compliant authorization. Microsoft stated its commitment to building 'Humanist Superintelligence' to ensure that cutting-edge AI serves as a tool for humans under human supervision. Additionally, Microsoft's GB200 computing cluster has been fully deployed to drive the continuous iteration of this model ecosystem. The flagship inference model of the MAI family, MAI-Thinking-1, features 35 billion active parameters, employs a mixture of experts (MoE) architecture, and provides a 128K context window. In mainstream software engineering and mathematical reasoning assessments such as SWE-bench Pro, this model achieved performance comparable to Claude Opus 4.6 and outperformed Claude Sonnet 4.6 in blind human evaluations. For programming scenarios, MAI introduced the MAI-Code-1-Flash, an agent programming model with 5 billion parameters, which will be deeply integrated into GitHub Copilot and VS Code, offering performance on par with Claude Haiku at a lower inference cost. In the multimodal domain, MAI-Image-2.5 and its Flash variant support high-precision text-to-image generation and image editing, with image quality scores surpassing Nano Banana Pro. For speech and transcription, MAI launched the MAI-Transcribe-1.5, a state-of-the-art transcription model for 43 languages with five times the speed of competitors, as well as the MAI-Voice-2 and its Flash variant, which supports 15 languages and features emotion control and zero-shot cloning. These models will not only be deployed on Azure AI Foundry but will also be available on OpenRouter, Fireworks, and Baseten, with developers being able to fine-tune weights for the first time. Microsoft also revealed a 1.4 times improvement in computational efficiency achieved through joint software-hardware optimization with its self-developed chip Maia 200. In addition to the basic model releases, Microsoft launched the 'Frontier Tuning' service based on a Reinforcement Learning Environment (RLE). This service allows enterprises to customize train MAI models using their internal operational trajectories, decision sequences, and specialized data in a fully controlled isolated environment ('training gym'). Tests indicate that custom models after Frontier Tuning significantly improve efficiency, with the MAI model optimized for Excel aligning with GPT-5.4 while being ten times more efficient. Furthermore, a custom MAI model for McKinsey achieved the highest win rate while reducing costs by nearly ten times. Additionally, Microsoft announced a strategic partnership with the Mayo Clinic, a leading global medical institution, to jointly develop a clinical reasoning large model based on Mayo's clinical data and Microsoft's AI foundation. This model will be owned by the Mayo Clinic and will first be deployed internally for early diagnosis and treatment planning, before being made available to other medical institutions through Azure AI Foundry.

Microsoft AI Launches First Inference Model MAI-Thinking-1 and Six New Native Models, Introduces Private Reinforcement Learning 'Frontier Tuning' Service

2 июня государственный секретарь США Рубио заявил, что получение ответа от Ирана может занять до 6 дней. (Цзиньши)

[Государственный секретарь США: Ответ Ирана может занять до 6 дней]2 июня государственный секретарь США Рубио заявил, что получение ответа от Ирана может занять до 6 дней. (Цзиньши)

Государственный секретарь США: Ответ Ирана может занять до 6 дней

Interview with Former Goldman Sachs CEO: Crisis, Leadership, AI, and Advice for Young People

2-Year Return of 225x? Uncovering Mysterious Researcher Serenity's AI 'Choke Point' Investment Strategy

Retail Investors' 'Lead Brother' Serenity vs. Newly Minted Stock God Leopold: How Are the Two Top Hunters Mining AI's 'Physical Limits'?

VVV Surges Over 10x This Year, Is the Base Ecosystem the Last Hope for Crypto AI?

Bitcoin Correction Pushes 580,000 BTC Into Loss Territory

One Article Deconstructs the Investment Methodology of 'Stock God Serenity'

Deconstructing the Investment Methodology of the 'Stock God Serenity' in One Article

Sharplink CEO: Ethereum's Future Is Playing Out Now

Kalshi and Coinbase Receive CFTC Approval, Ushering in the Most Regulation-Friendly Era for the Crypto Industry?

Weekly Token Unlocks: HYPE Unlocks $650M Worth of Tokens

Article

BTC falls below 70000 USDT

Strategy's First Sale Triggers Panic as Bitcoin Falls Below $70,000 Key Psychological Barrier

Государственный секретарь США: Ответ Ирана может занять до 6 дней

Disclaimer：The content above does not represent HTX's positions.，HTX does not provide any trading recommendations.。

All Comments0LatestHot

Trending Topics