released

IBM announced the next-generation superconducting quantum processor IBM Quantum Nighthawk r3, designed to accurately perform 10,000 quantum operations on 520 programmable qubits. This processor contains 2,034 physical qubits, continuing the square lattice connectivity and fast operation architecture of Nighthawk r2, and adds 10 inter-chip connections called l-couplers. IBM plans to connect two Nighthawk r3 processors into a single 1,040 qubit system by 2027.The currently deployed Nighthawk r2 is equipped with 120 programmable qubits, 218 couplers, and 120 reset elements, capable of executing over 100,000 circuits per second, with a throughput approximately 25 times that of Heron, and a maximum speed increase of 19 times for partner workloads, with E.ON reporting a 15 times increase. IBM previously connected two Nighthawk r1 processors using 4 l-couplers, achieving a quantum state transfer fidelity of up to 99.3%, averaging 99%; the two connected cryogenic modules cooled down to below 15 milli-Kelvin within 5 days.

Google released the Gemini Decisions API, with the underlying model being Gemini Decision 1. Sundar Pichai and Logan Kilpatrick announced this model. It handles closed-set decision-making, scoring options in a single forward pass rather than generating text token by token. The model supports yes/no judgments, multiple-choice, and scale ratings, with multiple-choice allowing up to 255 options, and can accept text, images, audio, video, and PDFs, with a maximum of 1000 questions per request, a context window of 256K, and an end-to-end latency of about 150 to 250 milliseconds, returning confidence for each answer.The input price is $0.10 per million tokens, with output not charged. Google states that this price is about one-fifth of Flash-Lite. The project began as an internal prototype at Google DeepMind, with a small team delivering the model and interface after about six weeks. Google’s self-testing shows that Decision 1 leads in 5 out of 7 shared benchmarks against the TypeSafe decision model Jev 1.14, with JevBench at 92.2% and Jev 1.14 at 90.0%. The input price for Jev is $0.042 per million tokens.

On November 12, at the opening ceremony of Baidu World 2026, Kunlun Chip, a subsidiary of Baidu, introduced its next product plans. The M100 has been launched as scheduled, and the M300 will be released next year.Among them, the M100 focuses on extreme cost performance; the flagship M300 has made comprehensive capability leaps in multiple dimensions such as computing power, storage, and connectivity. In addition, by 2028, multiple chips based on the new N architecture will also be launched to match differentiated application scenarios.Baidu founder Robin Li stated at the conference that the demand for AI computing power has shifted from being primarily training-focused to balancing both training and inference. The scaling of intelligent applications has driven a massive explosion in inference computing demand, "inference will become the main driving force for the long-term sustained growth of AI chips."

On November 12, at the Baidu World 2026 Conference, Baidu released the Tianchi 1024 card super node, which is planned to be launched next year. This super node uses optical engine technology to achieve full optical interconnection, supports collaboration of 1024 cards, with inter-card bandwidth increased by 3 times compared to the previous generation, training performance improved by 2.5 times, and inference performance enhanced by 3 times.The accompanying G3.5 KV Cache video memory acceleration cabinet can double the long context inference performance, reduce the first token waiting time by 90% in multi-turn dialogue scenarios, and increase overall throughput by 7 times. A single super node can support the training of large models with 50 trillion parameters, and can be scaled up to a cluster of 500,000 cards.This product was originally scheduled for release in 2028, but this announcement has moved the timeline forward by a year. On the same day, Baidu founder Robin Li revealed that the flagship Kunlun chip model M300 will also be launched next year.

ElizaOS founder Shaw stated that he previously referred to the code published in the GitHub repository lalalune/pumpfun-source as open source, but that statement is inaccurate. The relevant code does not belong to him, and he has never had the right to publish, license this code, or authorize anyone to use it. The "Copyright (c) 2026 Shaw" and similar licenses in the repository are not genuine licenses and do not grant anyone related rights. Shaw stated that his previous remarks or these licenses should not be regarded as permission to use, copy, fork, modify, develop based on, or distribute the code, nor should anyone distribute or attempt to reverse engineer the pump.fun code, as the relevant code is protected by copyright.

According to Tether's official announcement, Tether AI Research has released the open-weight model QVAC BotPsy, which has 800 million parameters and is based on the Alibaba Qwen3.5 architecture, aimed at tool invocation and agent tasks on personal devices.Tether stated that in four evaluations: BFCL v4, ACEBench, τ-bench, and ToolSandbox, BotPsy's scores are approximately twice that of its base model and surpass the 2 billion parameter Qwen3.5-2B-Think. The model can convert user requests into local tool calls, and developers can adapt and deploy its weights within their own systems.

According to IT Home, Baidu's Wenxin large model has released ERNIE 5.5 Flash Preview, which is the first post-training version based on the previously released Base model. This model has a total of 280 billion parameters, with 14 billion active parameters, and supports 1 million contexts. The official claims it performs excellently in programming, 3D modeling, knowledge work, content creation, and more.ERNIE 5.5 Flash Preview balances intelligence level and inference efficiency, achieving competitive scores in multiple benchmark tests while maintaining a fast decoding speed. The model API has been launched on the Qianfan platform, and the weights will be open-sourced before November 25.Baidu stated that this model is an intermediate version in the post-training process. Although it has achieved competitive performance among models of the same size, there may still be some capability deficiencies due to insufficient training. Post-training is still ongoing and will be optimized in the subsequent official version.

According to Cointelegraph, the Bitcoin Lightning Network wallet platform LNbits urgently released version 1.7.0 after experiencing a wave of attacks that led to stolen funds, urging users to update immediately. LNbits warned that the vulnerability affects all earlier versions, and simply upgrading will not protect users whose credentials have already been leaked. It is recommended to reset wallet keys, change relevant credentials, and migrate servers if an intrusion is suspected.LNbits founder Ben Arc stated that the team is working hard to respond after users and the team suffered losses. Some users reported that attackers created new accounts using their LNbits instances, stealing approximately 0.023 Bitcoin.

After the CPI data was released, the yield on U.S. Treasury bonds expanded its gains, with the latest 10-year Treasury yield rising by 9.03 basis points to 4.952%.The latest 2-year U.S. Treasury yield increased by 16.54 basis points to 4.715%.

According to Gate market data, U.S. stock index futures showed mixed trends after experiencing their best trading day, as the market awaits a key inflation report. S&P 500 futures rose slightly by less than 0.1%, while Nasdaq 100 futures fell by 0.4%. WTI crude oil prices dropped by 2%, to about $85 per barrel.

On November 10, the TRON ECO × BBW 2026 Thanksgiving Parade officially kicked off at its first stop, with five major projects: JUST, SUN, BitTorrent, WINkLink, and AINFT jointly offering 30 Binance Blockchain Week tickets and 5 ecological peripherals. From November 10 to 15, users can follow @DeFi_JUST, @OfficialSUNio, @BitTorrent, @WinkLink_Oracle, @AINFTcom, retweet the event tweet, and send the specified text in the comments, tag 3 friends, and use #TRONEcoBBW to participate.Participants must confirm their ability to attend BBW in Bangkok on November 28-29; winners must confirm within 48 hours after the list is announced, purchase their own tickets, and submit proof by November 19. After TRON ECO reviews and approves, the ticket costs that meet the requirements will be fully reimbursed according to the event rules. The first stop of the parade has set sail, and eligible users are welcome to participate and join us at the BBW venue in Bangkok.

According to Apple's security announcement, the iOS and iPadOS 27.2.1 update fixes multiple security vulnerabilities, one of which is a Messages vulnerability that could allow remote attackers to join iMessage group chats and receive messages.Another Messages vulnerability could allow new users joining a group chat to read messages sent before they joined. Apple also fixed issues where malicious Bluetooth devices could pair without user interaction, and where attackers within Bluetooth range could bypass pairing code attempt limits. Apple stated that the above update was released on November 9.

Meta released the model Muse Ember, positioning it as a small model aimed at complex reasoning execution and repetitive tasks for agents, suitable for tool invocation, coding sub-tasks, routing, information extraction, and long document parsing. This model natively supports text, images, audio, and video, capable of handling screenshots, charts, scanned documents, voice memos, meeting recordings, and video content.Muse Ember has been launched on Muse Code and Meta Model API, with the model identifier muse-ember-1.0, supporting the Responses API, a context window of 1 million tokens, and a maximum output of 512,000 tokens in a single instance. The pricing for prompts within 200,000 tokens is $0.10 per million input tokens, $0.004 for cached input, and $0.50 for output; for longer prompts, the corresponding prices are $0.30, $0.012, and $1.50 respectively.

The U.S. October CPI inflation report will be released at 21:30 Beijing time, but this report will lack the year-on-year data for CPI and core CPI.The reason is that the U.S. government shutdown in October 2025 prevented staff from collecting price data for that month, resulting in a lack of reference for this year. The U.S. Bureau of Labor Statistics is unable to calculate year-on-year data, and the market can only rely on month-on-month data to assess the inflation situation.Economists expect that the U.S. October 2026 CPI will rise by 0.3% month-on-month, and core CPI will rise by 0.2% month-on-month, with growth rates cooling down.

Anthropic released Claude Fable 5.2 and the restricted version Mythos 5.2 on Monday, stating that the new model can transform a small number of instructions into runnable software, design physical objects, and address research questions that previously required a professional team. The official announcement claims that its output speed has improved by up to 30% compared to the previous generation, with Fable 5.2 scoring 49.1% in the internal FrontierCode 2 evaluation, higher than Fable 5.1's 32.4%. Both share the same underlying model, with Fable having stricter protections against dangerous biological and cybersecurity requests, while Mythos is only open to certified institutions in relevant fields.In terms of pricing, the standard API price is $8 per million input tokens and $40 per million output tokens, lower than the previous generation's $10 and $50; an optional fast mode can increase output speed by up to 2.5 times, but the rates double to $16 and $80. In comparison, Kimi 3.1 Pro is priced at $2 and $10, while the latest DeepSeek model is $0.6 and $1.98 during off-peak hours.AI blogger Matthew Berman's tests show that Fable 5.2 can generate playable games and a browser version of Excel. One level styled after "Gears of War" took 43 hours and 48 minutes to complete, consuming 8.63 billion tokens, with an estimated cost of $14,710 based on developer pay-per-use estimates; the Excel clone is estimated at about $11,700, and a "Fall Guys" style game ran for 71 hours, consuming over 9 billion tokens, approximately $13,000. Anthropic provides one week of Fable access for Pro users and standard Team seats, capped at half of the weekly quota, with promotional access for Pro users ending on November 16.

The development team Sigma Prime of the Ethereum consensus client Lighthouse has released a high-priority security update and recommends node operators to complete the upgrade as soon as possible. The official statement indicates that specific details of the vulnerability will be disclosed after operators have had sufficient time to complete the upgrade.This update does not introduce any disruptive changes. Lighthouse is one of the mainstream clients of the Ethereum consensus layer, developed and maintained by the Sigma Prime team.

ByteDance Seed released the new generation image creation model Seedream 5.5 on November 9. This model focuses on upgrading three capabilities: achieving pixel-level preservation in non-editing areas during multi-round editing, eliminating color shifts, and reducing detail loss and image quality degradation during continuous modifications, supporting up to 100 rounds of continuous creation around the same work; natively outputting transparent or semi-transparent images with an Alpha channel, handling contour edges and transparent areas; while also improving lighting, material, and skin texture performance, reducing excessive glossiness and overexposure, and enhancing text rendering accuracy for complex glyphs and dense small text.Evaluation shows that Seedream 5.5 has improved generation effect scores and usability in six scenarios: e-commerce, marketing, film and television, gaming, office, and fashion design compared to Seedream 5.0 Pro, with the usability in film and office scenarios exceeding twice that of the previous version. The specialized evaluation for text structure and realism uses GSB scoring, covering text-to-image and image-to-image, with positive preferences in all four comparisons compared to the previous generation, and the realism image-to-image GSB is +55.62%. Compared to GPT Image 2.5 Sunburst and Nano Banana 2.1, Seedream 5.5 performs better in text structure and text-to-image realism. In evaluations covering both Chinese and English as well as dense small text, the severe text distortion rate has dropped to 0.4%, and the slight text distortion rate has dropped to 1.7%. Currently, this model is being gradually launched on Jidream, Doubao, Doubao Work, and Volcano Ark, with a Pro version to be released later.

According to Beating AI News, Alibaba has released Qianwen Office 2.0, which adds a "Trigger Task" feature. Previously, this product could only execute tasks automatically at preset times; now it can also allow AI to start processing automatically upon receiving specific messages, approvals, or pending events.For example, after the boss sends work arrangements via DingTalk, the AI can automatically organize the requests into to-do items according to pre-set rules. Users can view the execution results and can modify or turn off the rules at any time.Tasks can be executed on local computers or in the cloud. Pure cloud tasks do not rely on the computer being powered on and can continue running even after the computer is shut down.