Sakana upgrades its "Model Battle Team": Ultra v2 prioritizes performance, Max focuses on cost reduction.

Beating AI Express (Insight): AI Agent startup NeoCognition has released ApprenticeBench, a benchmark specifically designed to test whether AI can learn a job independently like a new hire. The first version places an Agent at a simulated California construction company to handle accounts payable: it reads six months of historical bills, company manuals, and Odoo tutorials, then processes 100 bills over seven consecutive months, with rule changes and supervisor feedback occurring throughout. Results show Fable 5.1 achieved a 72% success rate, GPT-6 Astra hit 68%, while the best human test subject scored just 51%. The task includes invoice entry, error detection, cost code allocation, vendor outreach, approval routing, and learning internal company rules from historical records and feedback. Notably, the two models performed slightly better when directly operating the GUI than when calling APIs, largely eliminating the performance loss common in Computer Use scenarios. However, the "surpassing humans" finding is limited to this single simulated role, with only two human test samples. Cost is also higher: Fable 5.1 averages $18.23 per task, compared to ~$7.21 for humans. Humans grow faster over time, while Agents slow down as they accumulate more stored information.

Well-known trader Killa has shared statistics showing that after the last three U.S. CPI releases, BTC has risen more than 5% within 8 days. Killa believes the market has already priced in negative news, and a bear trap has been set. “We’ll see if a similar pattern emerges this time.” As a BTC-focused quantitative trader, Killa accurately predicted the peak of the current bull market in May 2025 and boasts over 200,000 followers on X. In mid-April, he shorted Bitcoin at $74,688, then switched to a long position during the broad market sell-off on June 5.

Robinhood-linked crypto, stock, and meme coin issuance platform Pons announced that it has added new stock tokens SNOW, RCAT, SEWY, LMT, INTC, NET, BA, GLXY, RKLB, and CEG as base pools for its newly issued tokens. Stock Meme refers to an emerging concept combining traditional meme coins and tokenized US stocks: instead of pairing with USDT or ETH, meme coins now form trading pairs directly with on-chain US stock tokens (such as NVDA, TSLA, AAPL). This model retains meme coins’ high volatility and community-driven speculative attributes while leveraging the hype and narratives of real stocks. Additionally, a portion of transaction fees is returned to the community treasury to accumulate the corresponding US stock tokens, forming a dual-driven model of "sentiment speculation + real asset anchoring".

According to GMGN market data, the market capitalization of Commodity Market Exchange (CME), Robinhood’s on-chain token issuance platform, has exceeded $15 million, currently standing at $15.85 million, with a 14-fold increase in a single day. Public information shows that the platform supports token issuance paired with tokenized commodities as the underlying pool. BlockBeats reminds users that the relevant protocol is in its early stages, carries substantial uncertainty and high price volatility, so users should exercise caution with their investments.

According to Bitget's market data, the Nikkei 225 index closed down 1.93% to 64,011.34 points; South Korea's KOSPI index closed down 1.76% to 6,909.92 points. Among heavyweight stocks, SK Hynix fell 2.21% and Samsung dropped 3.53%.

According to TradingBeats monitoring, two HYPE short sellers on Hyperliquid hold a combined short position of approximately $40.971 million, with unrealized losses of around $7.964 million. Last night, the two placed 200 buy orders in the $64–$77.77 range, totaling about $33.953 million, waiting for the price to fall before covering their positions in batches. The address starting with 0x939f holds roughly 240,500 HYPE short contracts at an average price of $62.50, with unrealized losses of around $4.09 million. It placed 100 buy orders in the $70.1–$77.77 range worth $17.698 million, covering its entire short position. The other address starting with 0x37b8 holds about 274,800 HYPE short contracts with an average entry price of $65.41, and unrealized losses of approximately $3.875 million; it placed 100 buy orders in the $64–$72 range totaling $16.255 million, covering around 87.3% of its short position. Both sets of orders were primarily created last night, with some orders still adjusted this morning. HYPE is currently trading at around $79.51, just about 2.2% away from the nearest tier of $77.77 covering orders. If the price continues to drop to around $72, the second short seller’s large-scale covering plan will enter the execution range.