beta

The blockchain transaction dissemination infrastructure bloXroute announced that its Solana DEX aggregator blox has entered public testing. This product addresses changes in pool balances between quotes and execution, providing faster quote responses, and reassessing paths based on real-time liquidity during execution to optimize trade execution.In early mainnet testing, blox completed a total of 200,000 exchanges, with a trading volume of 25 million USD and 10,000 independent traders. A 7-day benchmark test showed that the median response time for its exchange interface was 9 milliseconds, with an average slippage of -0.2 basis points for SOL and USDC mainnet exchanges. During this period, all successful exchanges enabled dynamic routing, with 29% of paths completing active pool replacements, achieving a median gain of 97 basis points.The public testing interface is currently open for use, while the production interface requires a whitelist application to the bloXroute customer team.

OpenAI has released the beta version of Businesses in ChatGPT.

The mobile trading application beta of Leap, a subsidiary of Solana Labs, has been launched and is currently undergoing public testing via iOS TestFlight, with the Android version set to be released soon. The application is based on the Solana network and supports trading tokenized stocks, ETFs, precious metals, foreign exchange, and crypto assets, with spot trading priced in USD, one-click confirmation, and executed as an on-chain swap.Leap was developed by the Dialect team, which was acquired by Solana Labs earlier this year. This team previously developed Solana Actions, the Blinks protocol, and the cashtags feature for the X platform. The application integrates with Jupiter for trading routing, Helius for infrastructure, and Birdeye for market data; there are no fees charged during the initial launch phase, with plans to use fee revenue to buy and burn SOL in the future.

In the early morning of November 4, Apple opened testing for the latest iOS developer Beta version. According to Caixin, users can directly access the text and image understanding, content generation, and other capabilities of Alibaba's artificial intelligence application Qianwen through Siri, writing tools, and visual intelligence. Journalists found that users can complete basic Q&A, image-text analysis, and content generation on their iPhones, and further needs can be redirected to the Qianwen App, with conversation information automatically synchronized.Users can choose whether to enable the Qianwen integration feature. Once activated, they can log in using their Apple ID or their own Qianwen App account, and use Qianwen within Siri, visual intelligence, and writing tools. When journalists inquired about methods to improve sleep, Siri called Qianwen to provide general suggestions; for more personalized plans, users could click the Qianwen icon to jump to the Qianwen App, which analyzes influencing factors based on previous sleep records, exercise conditions, and lifestyle habits to formulate improvement plans.Visual intelligence supports the option "Ask Qianwen" to identify items or scenes in images, for example, first asking Siri about clothing brands, then jumping to the Qianwen App to find the same style on Taobao and place an order. The "Qianwen Creation" feature in the writing tool can generate content based on requirements and adjust tone and length, such as retaining the time and place in an event invitation or compressing it into a brief notification for easy forwarding. Currently, users in China who purchase iPhones that support Apple intelligence, or whose device location and Apple account country and region are both set to China, can experience related features through this developer Beta version.

The DeFi protocol Fluid has launched the DEX v2 beta version, which supports customizable collateral and debt ranges. Governance has set a borrowing limit of $50,000 for each asset for the new contracts deployed on Ethereum, while the creation of liquidity pools remains restricted.

According to "Du Jia," Tencent has launched the internal testing of "Tencent Cloud Drive," with some features still being iterated, and plans to release an App version. This product focuses on file storage, data management, and AI applications, with a core positioning as a unified memory base for multiple Agents, aiming to connect various AI applications and office tools, unify and carry context, allowing multiple Agents to collaborate based on the same information.Practical tests show that the internal test version retains basic entry points such as the homepage, files, recycle bin, and storage space, with a floating AI assistant panel on the right side. The core highlight of the product is its "connection" capability, supporting binding with Tencent Docs, WorkBuddy, and integrating third-party Agents like DeepSeek Harness and OpenClaw. After authorization, the Cloud Drive AI assistant can read application files, automatically sync updates, and directly call information as a knowledge base for retrieval, summarization, and content generation. The binding page will clearly inform users of the scope of permissions and the rules for AI credit consumption.The report points out that, unlike traditional Cloud Drive AI's document summarization, this system fully connects documents, Cloud Drive files, and Agent workspaces, shifting the Cloud Drive from a file repository to a data memory platform in the AI era.

According to "Du Jia," Tencent is currently beta testing "HermitKit," an AI capability plugin for WeChat Mini Programs, providing three levels of access: API, UI components, and Demo. This plugin supports streaming calls to large models through ChatApi, creates and manages Agents using AgentBuilder, and enables multi-turn conversations, with built-in session storage and long-term memory capabilities. Additionally, it can bind remote Agents via QR code scanning through AgentChannel to complete message sending and receiving.Currently, HermitKit showcases 7 types of functional demonstrations, including AI SDK, Agent Channel, Agent Management, Pet Agent, Chatbot, Travel Planning, and Plugin Market. Among them, the AI SDK aligns with cloud development APIs such as createModel, generateText, and streamText; the Pet Agent allows developers to read pet status through interfaces and trigger status changes using props; the Plugin Market enables searching, publishing, and downloading MCP, Skill, and Expert resource packages. This plugin has not officially launched yet, and related capabilities may still be subject to adjustments, with specifics to be confirmed by official statements.

According to The Defiant, Jupiter has launched a beta trading product called YOLO, which supports up to 1000 times leverage, and this product has not yet been audited.According to the official statement, users trade against the platform's capital pool (house) in this product. When the capital pool is insufficient, users' profits will be queued for processing, and there is no guaranteed payment time.

On October 8, Anthropic announced that Claude Dashboards and Claude Motion have entered beta testing. Docs, Slides, and Design have exited testing and are available in all Claude plans, including Free; since these features entered each Claude conversation, users have created over 45 million documents, presentations, and designs.Claude Dashboards is open for testing in paid plans and can connect to data platforms such as Amazon Redshift, BigQuery, ClickHouse, Databricks, Snowflake, and tools like Salesforce. After users ask questions in natural language, dashboards can be generated and continuously updated; clicking on numbers reveals corresponding queries, and charts display the last refresh time of the data. Dashboards can be sent to Amplitude, Grafana, Hex, Mixpanel, Omni, Perplexity, PostHog, Sigma for further analysis, with Looker, monday.com, and Tableau support coming soon.Claude Motion is entering testing in Team and Enterprise plans, allowing users to generate playable animated explanations based on prompts, supporting modifications in the editor or adjustments by Claude, and can be downloaded as MP4.

FastLane co-founder Alex Watts announced that the team is building Moose, a product that serves as an on-chain aggregator, which is currently in public testing on Monad. Moose is described as the first on-chain aggregator and is currently the only router aimed at addressing the malicious quote manipulation issues from prop AMM and v4 hooks in TradFi and DeFi.Its tech stack is inverted, with quotes being instant and private, not passing through servers, TEE, RFQ, or simulation in the hot path, and routing decisions made using on-chain real-time status after the trade is included. Moose provides improvements in execution, speed, spot margin safety, and privacy by co-locating with various DEXs within the same transaction. Gas on Monad is higher but below one cent, and the upcoming threshold encrypted memory pool will offer end-to-end privacy.

The core development team of Solana, Anza, posted on platform X stating that the mainnet Beta transaction V1 will begin activation at epoch 1035, expected around September 15 at 1:20 UTC.The ecosystem team indicated that more time is needed to test and integrate V1 support; V1 is currently live on the development network for integration testing. The computational budget in V1 is set through new transaction configuration settings; the computational budget instructions are still usable in traditional transactions and V0 transactions, but will not take effect in V1; application developers should check the impact of related changes on their applications even if they do not plan to send V1 transactions.

Anza stated that the reduction in Solana token rent has gone live on the mainnet Beta, and proposal SIMD-0437 Step 1 has been activated at epoch 1028.This step reduces the required lamports per byte from 6960 to 6333; subsequent functionality gates will be activated after checking the growth status at the current rent level, ultimately reducing to 696 gradually.

The Solana ecosystem DeFi protocol Kamino announced the launch of Kamino Fixed Rates, positioned as an on-chain fixed-term, fixed-rate credit architecture solution. The product is now live in a private testing version, with a public version planned for release in the third quarter of 2026.According to the introduction, Kamino Fixed Rates aims to provide a lending framework for the on-chain credit market with both the term and interest rate locked in advance. Kamino also released a related light paper for the market to review the product design and mechanism description.

Bitcoin infrastructure development company Blockstream announced on Twitter that due to Boltz's ongoing automated detection and multiple vulnerability attacks leading to the suspension of its exchange services, Blockstream will launch its own Blockstream Swaps service.Currently, Blockstream Swaps is in the beta testing phase, aimed at providing a toolkit for the ecosystem and will complement existing service providers.

According to blogger Max For AI, DeepSeek Harness (DSH) will officially start public testing on August 13. The final internal test version will be pushed tonight for developers to complete the final plugin compatibility work, and the plugin repository should include the #dsh topic tag. Plugins developed during the internal testing period can be migrated to the developer's personal account and made public. This means that DeepSeek is not only releasing the Harness product this time, but its accompanying plugin ecosystem will also be opened simultaneously. Currently, the official announcement has not been made, and specific details will be based on official information.

The Beta version of the tokenized asset on the Web3 asset data platform RootData is now online. This page aggregates and displays the market share of the tokenized stock market, popular activities, stock derivatives, and tokenized stock data based on publicly verifiable exchange quotes, order books, on-chain, and issuer data.It is reported that users can view the 24-hour trading volume, liquidity depth, open interest (OI), and the number of active underlying assets on platforms such as Binance, OKX, Gate, Bitget, Hyperliquid, and Bybit. Additionally, users can compare prices, spreads, depth, OI, and other indicators of stock derivatives across different platforms on a single stock detail page, as well as the trading volume, AUM, issuance, and circulation of the corresponding tokenized stocks. This page aims to help exchanges, issuers, and investors more intuitively track the popularity, liquidity distribution, and potential trading opportunities in the tokenized stock market.

According to Jinshi reports, the Chief Artificial Intelligence Officer of Meta Platforms announced that the Muse Code beta version has been launched. This is the first programming agent from the Meta Superintelligence Lab (MSL), built on Muse Spark 1.2.

The Beta version of the stock spot trading exchange ranking is now live on the Web3 asset data platform RootData. This ranking is based on the platform's publicly verifiable transaction and order book data (real stock trading data is not included), scoring and ranking based on nearly 24 hours of trading volume, liquidity, and the number of active stocks, helping users gain a deeper understanding of the current state of the stock market on various trading platforms and providing data support for further investment.According to this ranking, Binance, OKX, Gate, Bitget, and BingX are currently in the top five, with Binance having a trading volume of over $1.93 billion in the last 24 hours and a market share of 47.36%. In terms of the number of active products in the public market, Bitget, MEXC, and WEEX are in the top three. The exchanges and asset types covered by this ranking are still being further enriched.

DeepSeek releases update log, DeepSeek V4 Flash official version API is now live for public testing. The model structure, size, and DeepSeek V4 Flash preview of DeepSeek V4 Flash 0731 remain consistent, with only post-training being redone.The official version of DeepSeek V4 Pro will be released as soon as possible.

The B.AI platform announces that the Auto mode Beta version is now officially online. This mode does not require users to manually select a model; they only need to present their needs, and the system will automatically match the appropriate AI model based on the task type and generate responses, significantly reducing the cost of model selection.Key highlights include intelligent task matching, support for Web Search and tool capabilities, and no additional model selection fees, with usage being transparent and traceable. Users can now visit the B.AI official website to directly experience a new way of AI interaction with intelligent matching.