社群 · 3547

The Xiaomi MiMO-v3-Pro score is suspected to have been leaked, and the SW-Bench Pro reached 72.8 or close to the top overseas closed source model

According to Twitter news, Max For AI published an article that revealed that a Benchmark screenshot suspected to be Xiaomi's next big model, MiMO-v3-Pro, was circulating in the community. The screenshot shows that the model focuses on coding agents and general agent scenarios, and some test results are close to top overseas models such as Claude Opus and GPT. According to the suspected screenshot, MIMO-v3-Pro scored 72.8 points on the SW-Bench Pro, which is higher than the 67.9 points of GLM 5.3 and the 65.8 points of Kimi K3, which is less than 3 points different from Claude Opus 5's 74.6 and GPT-5.6 Sol Max's 75.4 points; Terminal-Bench 2.0 scored 70.6 points, which is also close to Claude Opus 5's 72.0 points and 73.5 points for GPT-5.6 Sol Max. Furthermore, it scored 76.4 points on the bt3-bench compared to the GPT-5.6 Sol Max with 78.8 points. If the above results are finally officially confirmed and replicated in the official version, the MIMO-v3-Pro may enter the first tier of the world's top models. However, at present, the authenticity and testing conditions of this Benchmark screenshot have not been officially confirmed by Xiaomi, and the relevant data should still be regarded as unconfirmed breaking news. It is worth noting that the latest MiMO flagships officially unveiled by Xiaomi are MiMO-v2-Pro and MiMO-v2.5-Pro, so whether V3-Pro exists and when it will be released is yet to be further disclosed by the official authorities.

1d ago

Bitcoin's rebound may just be a blood sacrifice for bears after bursting $1.1 billion overnight

Author: Shenchao TechFlow Original title: BTC's largest single-day short liquidation in history: $1.1 billion of short funds evaporated overnight, but shouting back is too early for every escape from death, requiring empty sacrifices. Bitcoin's intraday volume skyrocketed last night, once approaching the $70,000 mark. While the investment community is full of buoyancy, what you might not know is that last night was the largest single-day short liquidation in crypto history. The network's single-day bears rallied over 1.1 billion US dollars, breaking historical records in a fractured manner. The $1.1 billion bears were tightened overnight, breaking the crypto history record. On the night of August 19, BTC started around $64,000 and rose about 7% in an hour, reaching an intraday high of $69,970, just one step away from the 70,000 mark. This is the highest price since early June and the biggest one-day increase since March. The sharp rise was accompanied by blood washing in the contract market. According to public contract data, the entire network closed out about US$1,345 million in the past 24 hours, involving 105,000 traders, of which short orders were about US$1,191 million, and the long ones were only 153 million. In the most intense hour, the entire network sold out $1,194 million, with bears accounting for 93.5%. By currency type, Bitcoin contract bears were liquidated by about $662 million in 24 hours, while Ethereum bears were about $366 million. Leveraged positions that bet on falling were uprooted almost at the same time. Several whale positions on Hyperliquid totaling nearly $200 million (large highly leveraged companies) have also been completely liquidated. The liquidation itself will speed up the market. A strong bearish position means being forced to buy back up. The higher the price, the more explosive the higher the higher the higher the higher the higher the higher the higher the higher the price, forming a self-reinforcing feedback loop. At the same time, judging from multiple data sources, this is the largest Bitcoin shorting settlement in a single day. The White House summit was only the trigger; the catalyst came from the bond market attributing the surge to the combined benefits of the two. One is the message side. On August 19, Trump met with crypto industry executives such as Coinbase, Kraken's parent company Payward, and Blockchain.com at the White House, and the market's optimism about the shift in regulation heats up. The other one is lower level. On the same day, the US Treasury Department announced a direct doubling of the liquidity support for long-term treasury bond repurchases, raising the upper limit of a single operation from 2 billion US dollars to at least 4 billion US dollars, effective September 9. In the crypto community's view, this is a more tangible sign than the summit: macro liquidity is loosening in the direction of risky assets. The data also confirms institutional buying. U.S. spot Bitcoin ETFs had a net inflow of $297.6 million on Monday and another $189 million on Tuesday; funding rates have risen to a 20-month high. The bulls are crowded, and the bears are even more crowded. Once the price starts, a strong pedal will automatically be relayed. The last big bear liquidation dates back to 5.19, and old chives must remember “5.19” in May 2021: China clearly prohibited financial institutions and payment institutions from carrying out virtual currency-related business. Amidst the panic, Bitcoin hit more than 40,000 US dollars to around 30,000 US dollars in one day, and the entire network closed out about 7 billion US dollars in 24 hours, setting a historical record at the time. The 5.19 collapse liquidated bulls, and about $7.56 billion in long leveraged positions were instantly washed away. However, over the next few days, in the midst of panic, a large number of traders frantically leveraged at the bottom to chase the sky. As a result, there was an extremely violent retaliatory backlash. According to K33 Research quoting Coinglass data, short positions of approximately $757 million in a single day were instantly washed away, making it the largest day for BTC perpetual short settlements in history. And that record was broken just yesterday. After the liquidation of top bears, the market is often polished for a few weeks, and the liquidation of top bears is often a sign that the mid-term phased bottom has been completely consolidated. Positions are cleared after deleveraging, and macro-liquidity is transferred to encryption, and it takes weeks of cold washing and energy in the middle. Will this time be the same? Bitcoin is currently around $69,200, and the 24-hour increase narrowed to 7.6% (as of the morning of August 20). Sentiment indicators are still hesitating: the fear and greed index is 46, which is still in the fear zone; the probability that Bitcoin will reach 70,000 this month on the Polymarket has jumped to about 70% from before the surge. Axel Rudolph, IG's chief technical analyst, observes: Bitcoin is moving towards $70,000, driven by bears' recovery, indicating that buyers are...

2d agoburnking#Bitcoin

Why is capital chasing AI Native and ignoring the old Internet

Capital doesn't reward being old-fashioned, not because old-fashioned people are at fault. The old part is clearly priced. There is no bad information, so there is no excess profit. Global venture capital was $510 billion in the first half of 2026, surpassing $44 billion for the full year of 2025 in one and a half months. More than 70% have entered AI; OpenAI and Anthropic took 217 billion dollars, accounting for 43%. With that much money, you'd think everyone could share a little bit. The truth is that distribution is more extreme than total volume, and the first sieve doesn't screen the industry, it screens people. The category that has been screened out now has an unkind name: the internet is old. Let's just say one thing: the “old man” in this article has nothing to do with age. It refers to a set of methodologies that have been formed in the mobile internet cycle, have been tested over and over, and have brought huge returns to holders. The person holding it may be 45 years old or 32 years old. It was this methodology that was being repriced, not the year of birth. Confusing these two things is Lao Deng's most common mistake and one of the most comfortable mistakes — because if the problem is someone else's age discrimination, you don't need to change a single word. 01 What is AI Native The term has been misused. They can use ChatGPT not called AI native, nor AI in the company name, let alone in their twenties. There are three things that really separate people. First, the starting point is a model, not a requirement. The order in which Lao Deng makes a product is: look at what the user wants, write down the requirements, and find technology to implement it. The order of AI natives is reversed: first figure out what level the model is capable of today and what step it is likely to reach tomorrow, and then move from this capability boundary to the external product. The former uses the model as a tool, and the latter uses the model as the foundation. There was no difference between these two kinds of things made by humans in the first edition; by the third edition, there was a difference of one species. Article 2. The default unit of an organization is not a person. The division of labor in the Internet age is the division of one thing into ten people. AI Native's division of labor is to take ten things from one person and add a bunch of agents. The CEO of a domestic application company said that the team consists of less than ten people, but a large number of AI work at night, and the first thing employees do every morning is check the work the AI handed in the night before. Cursor's side is even more extreme. Public reports mention that the company doesn't have a product manager; engineers write their own code, talk to users themselves, and participate in recruiting people themselves. Article 3. Information is first-hand. AI Native's input sources are papers, model cards, GitHub issues, original discussions on X, and self-run evals. Lao Deng's input sources are industry summits, closed-door meetings, brokerage reports, interpretation of public accounts, and finding someone to drink coffee with. This one is the least obscure and most lethal; I'll talk about that separately later. I'm satisfied with all three. The 25-year-old is an AI native, and so is the 45-year-old. I'm not satisfied with the three rules; I'm still an old man at the age of 25. AI natives are a state, not an age group. The trouble is that tickets in this state are works, not resumes. 02 The two lists spread the results of this round on the table. These are two lists. The first one is an all-AI native company. Their valuations are not rising; they are exchanging orders of magnitude. List 1 · Upstream OpenAI raised $122 billion in a single round of financing in Q1 2026, followed by $852 billion, the largest private equity financing in history. Anthropic Q2 had a single round of $65 billion, after investing $965 billion, accounting for about half of the total global venture capital for the quarter; the revenue operating rate in May reached about $47 billion. DeepSeek raised about 70 billion yuan in its first round of financing in May 2026. In April of the same year, Liang Wenfeng raised his direct shareholding from 1% to 34%, and controlled a total of about 84.29% of the shares through related entities. The Dark Side of the Moon (Kimi) was estimated at $4.3 billion in December 2025; it went for three consecutive rounds from January to February 2026 to reach 18 billion; the D round in May was about $2 billion, breaking 20 billion dollars after the investment; the July round surpassed $3.5 billion, after investing 35 billion dollars; the pre-IPO target was 50 billion dollars. ARR broke 100 million in March, 200 million in May, and held steady at 300 million US dollars in June, with APIs accounting for more than 70%. Smart Spectrum · MiniMax successively landed in Hong Kong stocks in early 2026, with a market capitalization exceeding 100 billion yuan. It was one of the first major model companies listed in China. The second one...

3d agoWendy#AI #DeepSeek

Meme Express: The utility community is hotly debated on BSC, with a market value of $780,000

Comparing the news, the current market value is 780,000 US dollars, which is currently being discussed in the utility community on BSC. CZ posted an article on January 30 mentioning “GME should issue a utility token on the blockchain, on #BSC hopefully.” Multiple users on X said that the CA was GME's “utility” version and new story on BSC, and called it “Dragon 2 $UTILITY” compared to GME “Dragon 1.” This article is sponsored by GENG, Build Your Fortune on GENG (https://geng.one)

10d agoburnking

Some people use it, have brands, and don't issue coins. Why has POAP reached the end?

Author: imToken Original title: When POAP also comes to an end: When the “wave of bankruptcies” in the crypto industry hits, how can ordinary users cope with themselves? Recently, the crypto industry seems to have entered an intensive farewell period. From BitMEX, which has been in operation for 11 years, to Satori Finance, which has received investment from top institutions such as Polychain and Coinbase Venture, one familiar name after another has ceased operations and officially reached the end, covering various directions such as trading platforms, DeFi, wallets, NFTs, and infrastructure. Among them, POAP's departure was unquestionably particularly impressive. If you've gone through the previous crypto cycle, especially if you've participated in Devcon, ETHDenver, Hackathon, DAO community events, or various online and offline meetups, many people can probably pull out a few POAPs from their wallets. It may be from a conference, an online sharing, or just a community event where you can't remember the details. Most of these POAPs aren't worth much, but because of this, they're probably closer to the original meaning of “collecting” than many NFTs that used to be expensive. It is for this reason that POAP's farewell is particularly representative. It didn't suddenly go back to zero due to hacker attacks, and didn't even issue a native token that needed to continuously maintain price expectations. It just had real users, clear scenarios, and a high enough brand awareness, but in the end, it still hasn't found a business model that can support the company for a long time. This is exactly what is changing in the crypto industry today. In the past, we were more accustomed to discussing how a project was born; next, we may need to get used to discussing how a project dies. And this isn't necessarily a bad thing. However, as regular users, we need to know how to avoid being affected by the aftershocks of a bear market. 1. A new form of “shutdown wave” swept through the Web3 encryption industry. In the last round of expansion, it is actually not difficult for a project to prove that it was “founded.” The completion of financing, the launch of the main network, the issuance/airdrop, and a round of liquidity incentives are enough to attract the first batch of users. TVL, number of addresses, and transaction volume can quickly grow. Even over a long period of time, whether a project actually has revenue is not the most urgent issue. However, when the cycle is reversed, and token prices and liquidity cannot continue to perform financing functions, this model will reveal one of the easiest questions, which is, if no new money comes in, can this project support itself? This round of projects came to an end in 2026, and this is where the real focus is also on. Because many of those that have disappeared are not air projects that had no products at the beginning, but projects that have already been funded, launched, have real users, and even run well technically. For example, on July 23, BitMEX announced that it would officially shut down the trading platform on September 23, 2026. This trading platform, founded in 2014, was once one of the most representative companies in the entire crypto derivatives market. Perpetual contracts, 100x leverage, and a complete set of trading products that were later widely used by the entire industry are closely related to the early development of BitMEX. It even specifically emphasized in its official shutdown announcement that “in more than 11 years of operation, BitMEX has never lost user funds due to hacking,” but this has not made it an infrastructure that can run permanently. A similar story happened on the DeFi and infrastructure circuit. As a Bitcoin L2 project that has been under construction for nearly four years, Botanix has maintained 100% normal operation and zero security incidents since its launch. It has processed about 25 million transactions, 200,000 wallet addresses, and tens of millions of dollars of assets have entered the network, and is connected to infrastructure and DeFi products such as Chainlink and Morpho. Looking only at traditional Crypto KPIs, it's hard to even call it a “no-go” project — the chain has been created, the products can be used, the users have come, and the money has come in. But in the end, Botanix decided to shut down the network, and the review showed that the actual transaction demand was insufficient to generate sufficient fee revenue and could not cover the infrastructure costs required for the long-term operation of an independent network. At the end of the day, Crypto used to be too used to measure an ecosystem with TVL, number of addresses, and number of transactions, but it rarely asked that last question:...

12d agoburnking

SK Hynix labor conflict escalates again: employees plan to set up new trade unions or influence plans to expand production

Comparatively, SK Hynix labor and management continue to be at an impasse over the plan to distribute performance bonuses in the form of company shares. Some employees are promoting the establishment of a new unified trade union. Currently, a “unified trade union” preparation team has been set up. The relevant community has attracted about 4,000 employees to join, accounting for 11.6% of its total workforce. Market participants pointed out that if the labor dispute spreads to the production line, it may increase the uncertainty of SK Hynix's HBM production capacity expansion in the second half of the year. SK Hynix began mass production and shipment of HBM4 in the second quarter of this year, and plans to further expand production in the second half of this year. Currently, production has not been affected, but if the conflict escalates and spreads to manufacturing sites in the future, it may put pressure on the supply plans of major customers such as Nvidia. (Yonhap)

12d ago

He didn't understand the ball and relied on group chat. He became number one in the Moment prediction list

How complicated is it to predict a match? Research professional models, keep an eye on the game, invest money, or are you familiar with every player's data? For Asep Beton, the answer isn't that complicated. He's not usually a veteran soccer fan, and he doesn't track the news of all leagues and players every day. However, in the World Cup Call Challenge World Cup prediction event hosted by Moment, he checked the data, read game data, and discussed the team lineup and player status with his friends in the group. In the end, he won with the top score of the 11,412 event users and received the highest reward. There are no complicated trading operations, and there is no need to invest capital first. Users only need to choose between the two results based on their own judgment. This is where Moment's interest is most likely: turning predictions into something simple, intuitive, and judgmental. Not usually a die-hard fan, but was rewarded in World Cup predictions. In an interview with Moment, Asep honestly stated that he doesn't usually pay particularly close attention to soccer. What really gave him confidence to participate in the World Cup prediction campaign was a private chat group made up of friends. Before the game starts, everyone will discuss the team formation, players' recent performance, historical game data, and possible tactics in the group. Sometimes they share a piece of team news, sometimes they analyze whether a key player can play, and others make completely different game judgments. Asep then synthesizes this information to make its final choice. It doesn't sound mysterious, and it's not a professional strategy that's hard to replicate. Many regular fans usually talk to their friends about “who is more likely to win” and “whether there will be a certain result this time”. What Moment does is turn this everyday discussion into a predictive scenario you can actually participate in. Watch the data, talk about the competition, make choices, and wait for the results to be announced. It was through this method that Asep made correct judgments many times during the World Cup prediction campaign, and finally achieved first place in the standings. How to play Moment? In fact, it is a simple binary prediction. Many people have heard of “prediction platforms” for the first time. They may feel that the rules are complicated, or even think of professional trading. But Moment's core gameplay is pretty intuitive: binary predictions. Faced with a predictive event, the platform will give two clear options, and users will choose one of them based on their own judgment. You don't need to understand complicated K lines or master complicated trading tools. The engagement process can be simply summarized in three steps: read the prediction questions; choose what you think is more likely to happen; and wait for the event to end and complete the settlement. For users familiar with relevant events, it is possible to make judgments based on news, data, and experience; for new users new to forecasting, it is also possible to start experimenting with soccer, sports, or hot topics that interest them. There are only two choices, but the judging process can be very interesting. Some people rely on statistics, others focus on live lineups, some believe in long-term experience, and others, like Asep, first exchange different opinions with friends and then make their own decisions. There is no need to invest in the event, and you can participate in the competition by relying on judgment. Asep initially decided to participate in the World Cup Call Challenge. The most immediate reason is that the event rewards are attractive. But what really made him recognize Moment was that participating in the World Cup did not require a capital investment. Users don't have to deposit funds first, and they don't need to invest more to get higher rankings. Everyone can participate in predictions under relatively lightweight conditions, accumulate points through their own research, judgment, and prediction results, and compete for activity rewards. For those new to predictive gameplay, this model lowers the threshold of experience. Users can first familiarize themselves with the platform rules, observe the results of different events, and test their judgment ability without facing financial pressure from the beginning. This also brought the focus of the activity from “how much was invested” to “whether the judgment was accurate.” In an interview, Asep said that what he likes most is this: the event rewards analysis and predictive performance, rather than requiring users to invest a lot of money first. The rewards attracted him to participate. The simple gameplay made him willing to continue participating in Asep's experience with other prediction market platforms, but due to the relevant regulations in his country and region, he did not continue to use it later. In contrast, he thinks Moment's previous World Cup event experience was easier. You can participate in the competition by researching contests and making predictions without investing any capital. Throughout the process, what he enjoyed the most was not simply waiting for the lottery results,...

18d agoWendy#Moment

Well-known trader CryptoD: The bottom layout window before the next round of the bull market may appear in Q3 and Q4

Comparing the news, the well-known trader CryptoD posted on social media that looking back on social media recently, he found that the entire meme ecosystem has become very fragmented. The Chinese and foreign language markets are gradually forming different ecosystems, making it impossible to concentrate capital effectively. In addition, the market has been bearish for a long time, and there has been no significant increase in capital. Instead, there are more and more chains, and liquidity is becoming more and more scattered. However, I also think that now that the overall crypto market is gradually approaching the bottom, Q3 and Q4 will probably be the last period of time to calmly lay out before the next round of the bull market. CryptoD made a profit of over $10 million by trading the meme coin TRUMP alone.

19d ago

Bonk Guy: Money blindly pursues a new chain meme narrative, most projects lack a real user base

Comparing the news, crypto KOL Unipcs (also known as “Bonk Guy”) posted an article stating that the trending story of Crypto Twitter (CT) and on-chain traders continuously rotating to chase each newly launched public chain is essentially a continuous search for new loss opportunities. Unipcs said that two weeks ago, the market focused on hyping up the Base ecological narrative due to the performance of Robinhood Chain, but in the end, the market did not continue. Last week, the market began to see Stable Chain as the next “Robinhood Chain” opportunity, and he himself had previously alerted the community to related risks. He said that currently the meme coin with the highest listing value on the Stable Chain has dropped by about 90%, and most meme projects have not received real market attention. Unipcs believes that it is not opposed to Tether or Arc Chain, both of which are likely to develop if they focus on stablecoin infrastructure positioning. However, forcibly incorporating every new chain into the meme coin narrative is an irrational market act. He said that many participants are unaware of the meme coin culture, which could lead to a sharp decline in investors' assets. For trading meme coins, Unipcs suggests focusing on ecologically based networks such as Robinhood Chain, Solana, and BNB Chain. This article is sponsored by GENG, Build Your Fortune on GENG (https://geng.one)

22d agoburnking

KOSPI melted seven times during the year, and global stock markets are replicating the coin industry script

Author: Doumaru Original title: Welcome back to your native family. The global stock market is being coin-circled in Seoul on July 13, 2026. Korea Composite Index KOSPI plummeted 8.95% in a single day, breaking for the 7th time in the year. SK Hynix, the “national transport stock” in the eyes of Koreans, fell sharply by 15.37% in a single day. We have not seen such a sharp decline in nearly 20 years. Samsung Electronics also dropped more than 10%. More than 1.2 million leveraged accounts popped up with deposit recovery notifications, and the brokerage system automatically closed 320,000 to 460,000 accounts. What's even more worrisome is that 62% of those who closed out positions were young people aged 20 to 30, some lost their down payment for a wedding house, and others took loans to trade stocks... A young man in his 20s in Busan directly stabbed the blogger because he listened to the loss recommended by a stock YouTuber. In the past, these words were probably used to describe the scene after the coin industry plummeted; now, it is being repeated in the South Korean, American, and Japanese markets after the decline in technology stocks. The sharp rise and fall is only an indication. What really changed was the pricing method. The narrative overweighed the valuation, the leverage amplified sentiment, and social media quickly pushed the consensus to the extreme. The global stock market, especially technology stocks, is becoming more and more like a coin. Return to Native Family “Welcome back to Native Family.” After the sharp decline, traders in the coin industry who switched to the stock market wrote short loss essays one after another. The above comments can be found everywhere in the comment area. The so-called “native family” refers to cryptocurrencies. From the second half of 2025 to the beginning of 2026, a “run away from the native family” drama was staged in the coin industry. A group of KOLs and veteran players who have been struggling in the cryptocurrency market for many years are beginning to lose confidence in the cryptocurrency industry. Bitcoin fluctuated sideways, trading volume was weak, and meme coins were cut back and forth. Many people felt that “this circle is weak” and began to turn their attention to US stocks. This seems like a reasonable choice. Shares have revenue, profit, financial reports, and are regulated by the SEC. Compared to crypto projects that lack cash flow and rely entirely on consensus pricing, US stocks are at least a more mature and secure asset. Not only did traders in the coin industry take away liquidity, but they also carried the trading methods of the past. In the crypto market, they are used to chasing new narratives, finding highly flexible targets, using leverage, and quickly changing positions based on social media sentiment. Since entering the stock market, this method has hardly changed, except that trading partners have switched from tokens to AI, memory chips, and leveraged ETFs, and achieved great results over and over again. Storage stocks soon became the new collective consensus. The logic is not complicated: AI servers require more high-bandwidth memory, HBM is in short supply, and storage prices are rising. Naturally, Micron, Samsung Electronics, and SK Hynix have become the most direct “sellers”. Brother Sun's phrase “never lacking storage” is even more deeply rooted in the hearts of the people. Many crypto KOLs have changed and started talking about US stocks, storage cycles, and AI capital expenses. Doubling products such as SK Hynix is also used as a “more efficient” betting tool than ordinary stocks. Until the market reversed in July. Instead, Bitcoin has become a “low volatility asset”. How long will it take to drop in half from the top? It took 268 days for Bitcoin and 169 days for silver to complete a similar retracement. In contrast, it only took 36 days for SanDisk to drop about 55%, and it only took 34 days for SK Hynix to drop about 53%. Also, it was a “backlog”. Bitcoin was used for almost nine months, and storage stocks only used for over a month. This is exactly what is strange about this round of the market: in the past, investors feared that Bitcoin would soar and fall within a few days, while stocks relied on profits and slow valuation adjustments; now, some tech stocks are completing a complete bubble burst in a shorter period of time than cryptocurrencies. Very counterintuitive. Compared to some tech stocks, Bitcoin is becoming relatively stable. According to Carson Wealth Management statistics, Bitcoin's historical volatility in 2025 was about 42%, with a maximum retracement of about 32%; during the same period, Tesla's volatility was about 63%, the maximum retracement was 48%, Nvidia's volatility was about 50%, and the maximum retracement was 37%. Bitcoin is still a high-risk asset, but some big tech stocks are more volatile. In its 2026 outlook, Bitwise even predicted that Bitcoin's overall volatility may continue to be lower than Nvidia. So the current situation is quite absurd: Bitcoin is becoming more and more like technology stocks, but technology stocks are becoming more and more like Bitcoin. There is an old saying in the coin industry when narratives become valuation anchors; speculating on coins is speculating on narratives. Global tech stocks in 2026 are turning this statement into reality. AI is certainly not air, Nvidia...

23d agoburnking#KOSPI #stocks