According to DB News, the DOJ is cleared to sell 69,370 $BTC($6.58B) confiscated from Silk Road.

Beating AI News Flash: Firecrawl secures $75 million Series B funding led by Smash Capital, and launches Alexandria. The platform integrates web search, third-party data sources, and Firecrawl’s proprietary index into a unified interface, eliminating the need for AI agents to connect to separate APIs for each data type. For instance, an agent researching a company can first search the web, then directly call Fiscal.ai for financial data; access economic data via FRED and the World Bank; and search GitHub and Firecrawl’s Developer Index for coding-related queries. Alexandria currently covers additional data types including academic papers, government datasets, commodities, real estate, and job postings, accessible via API, MCP, and CLI. Firecrawl reported that when testing the same model and prompt across 845 tasks, responses using Alexandria were 21% higher in quality than those generated by the model’s native web tools. Proceeds from the funding round will also be used to acquire and integrate additional data sources. Firecrawl already has a paid partnership with Wikimedia Enterprise, and plans to onboard more researchers, publishers, and data providers to contribute their content to Alexandria, with revenue generated when agents utilize this data.

According to Bitget market data, South Korea's KOSPI index closed up 61.02 points, a 0.87% gain, ending at 7078.93 points. Samsung Electronics rose 3.25%, and SK Hynix gained 1.2%.

Official data from Hyperliquid shows its open interest has reached $18 billion, hitting a new all-time high. This figure refers to bilateral open interest, calculated as the sum of long and short positions. Over the past year, Hyperliquid has expanded from a crypto perpetual trading platform into an on-chain trading infrastructure supporting stocks, commodities, indices, pre-IPO, and event contracts. Launched in October 2025, HIP-3 allows third parties to stake HYPE and deploy their own perpetual markets, with underlying assets covering U.S. stocks, gold, crude oil, the S&P 500, and private companies like SpaceX. As of early September 2026, cumulative trading volume on HIP-3 markets exceeded $548 billion, accounting for roughly 30% of the platform’s total trading volume in the past 30 days. HIP-4, launched in May 2026, filled the gap in prediction markets and event contracts, and was further opened to third-party deployments at the end of August. During the same period, firms including 21Shares, Bitwise, and Grayscale sequentially launched HYPE spot or staking-based ETFs, while the platform rolled out native lending and advanced licensed HIP-3 markets for U.S. users.

Stablecoin payment network Atum has announced the completion of a $13.5 million funding round, led by Variant, with participation from PayPal Ventures, Abstract Ventures, Road Capital, Mirana Ventures, First Commit, and Credibly Neutral. Founded by former Visa crypto product head Pete Cooling, Atum positions itself as a unified coordination layer for global fund flows. The platform does not issue its own token, operate a dedicated blockchain, provide asset custody, or favor specific payment rails. Instead, it uses a matching market to let settlement service providers compete to fulfill payment requests, supports senders to freely choose payment assets and receivers to accept settlement assets as needed, and enables stablecoin micropayments and cross-border settlements for both humans and AI agents.

Beating AI Express News: Artificial Analysis has completed third-party testing of several new AI models released overnight. Claude Opus 5.5 max scored 58 points on the Intelligence Index v4.3, currently ranking first. GPT-6 Astra and Fable 5.1 both scored 53 points, while Opus 5 scored 51 points. Opus 5.5 max took the top score in 6 of the 10 tests. However, achieving these top scores required more tokens: Opus 5.5 max generates an average of around 119,000 tokens per task, roughly 60% more than Opus 5’s 73,000 tokens per task. Thanks to a 20% API price cut and 60% reduction in cache read costs, its per-task cost came to $5.98, almost on par with Opus 5’s $5.86. The more practical medium-tier model scored 51 points, matching Opus 5 max, with a per-task cost of just $1.34. OpenAI took a different approach: GPT-6 Sol and Luna delivered Intelligence Index performance close to their GPT-5.6 predecessor, but at significantly lower prices. Sol max’s per-task cost dropped from $1.99 to $1.06, while Luna’s fell from $0.18 to $0.07. In coding agent tests, Sol’s score rose from 55 to 57 points, with its cost cut by roughly half; Luna’s score dipped from 43 to 41 points, but its per-task cost is about 60% lower. The two companies’ strategies are starting to diverge: Opus 5.5 pushes performance limits higher with more inference, while GPT-6 Sol and Luna focus on making existing capabilities cheaper. On Artificial Analysis’s performance-cost chart, both companies’ new models occupy the new Pareto frontier.

Per Bitget market data, Hong Kong-listed AI application stocks extended their afternoon decline: Zhipu fell over 10%, Kingdee International dropped more than 7%, while Meitu, MINIMAX-W and other related stocks also tracked lower.