English text is being prepared.

English translation is being prepared.

Google Releases Gemini 4 Carbon, Claims SOTA on Cybersecurity and Long-Horizon Tasks

Google has released Gemini 4 Carbon, claiming it is a major improvement over Argon, which was released two weeks ago. The model achieves SOTA on cybersecurity and long-horizon tasks, ranks No. 1 on Artificial Analysis V5, shows significantly improved usability in areas such as coding, and is more closely aligned.

Vercel CEO Explains Vercel CI in Detail: Infrastructure, Turborepo Integration, and Agentic CI

Vercel CEO Guillermo Rauch published a post explaining how Vercel CI solves the AI coding bottleneck, in three layers: fast and stable infrastructure (20 clusters worldwide, running millions of microVMs per day), Turborepo integration (automatically computing the dependency graph and skipping and caching existing work), and Agentic CI. He said that vercel ci in the CLI is the first step; agents can now inspect, abort, trigger, and replay CI, and Vercel Agent can assist with optimization. More capabilities will be added later, giving CI engineer-like judgment. The official Vercel account previously announced that Vercel CI had entered beta, with CI in internal repositories sped up 2-6×, and optimized for agents via Vercel MCP and the CLI.

Google releases Gemini 4 Carbon, claims top ranking on Artificial Analysis V5

Google has released a new model, Gemini 4 Carbon, claiming it ranks first on Artificial Analysis V5 and is being made available first to cybersecurity defenders. The model shows notable improvements on coding and agentic tasks, scoring 88.2% on SEC-bench Pro vulnerability discovery, 54.8% on ExploitGym exploitation, and 71.0% on CWE-bench patching. Logan Kilpatrick said it is a big jump over Argon, which was released just two weeks ago, achieves SOTA on long-horizon tasks, and shows improvements in usability and alignment.

Mistral releases Mistral Large 4 (Le Chonk) weights on Hugging Face

Mistral released the model weights of the new version of Le Chonk (Mistral Large 4) on Hugging Face, targeting agentic coding and cybersecurity scenarios. Mistral claims that the model ranks first among open-source models on Cognition's FrontierCode benchmark and sets a new benchmark for Pareto efficiency on Artificial Analysis, while also improving inference efficiency and reducing inference costs. The model is also claimed to achieve SOTA in cybersecurity defense and has been tested by vendors such as CrowdStrike, Cloudflare, Rubrik, and Zscaler; users can customize and self-deploy it with Mistral Forge.

Mistral AI opens the Le Chonk model weights

Mistral AI announced the open release of the model weights for a new version of Le Chonk, saying it advances the open frontier in agentic coding and cybersecurity. The model ranks first among open models on Cognition's FrontierCode and sets a new benchmark for Pareto efficiency on Artificial Analysis; on cybersecurity it claims state-of-the-art, having been tested by vendors including CrowdStrike, Cloudflare, Rubrik, and Zscaler. The weights can be downloaded on Hugging Face and customized and deployed with Mistral Forge. Arthur Mensch reposted and said progress over recent weeks has been good, and now the weights are open.

Google Releases Gemini 4 Carbon, Focused on Cybersecurity and Long-Horizon Task Capabilities

Google announced the launch of Gemini 4 Carbon, calling it its strongest model to date and saying it reaches leading levels in cybersecurity and long-horizon task capabilities. Carbon improves on its predecessor on most benchmarks and stands out in autonomous coding and cyber defense. The model is first being made available to some trusted cyber defenders in the Fairwind program, with access to be expanded as soon as possible afterward.

Coforge launches Autonomous Software Factory to orchestrate AI agents across the software development lifecycle

Coforge has launched the Autonomous Software Factory (ASF) to orchestrate AI agents across the software development lifecycle, covering coding, testing, validation, code review, and quality assurance, with specialized agents able to execute these tasks in parallel within secure execution environments. Coforge says the model includes governance and human oversight, but did not disclose implementation details, customer deployments, commercial terms, or quantified productivity gains. HCLTech also plans to integrate Anthropic's Claude and Claude Code into its AI Force platform.

NVIDIA Releases Nemotron 3.5 Transcribe Open-Weight Speech Model, Supporting 40 Languages

On October 28, NVIDIA released Nemotron 3.5 Transcribe, an open-weight speech model that can transcribe audio in 40 languages, translate speech, and identify speakers, and developers can control the transcription format with natural language instructions without retraining. The model consists of a Transformer speech encoder, a speaker diarization model, and the Nemotron 3.5 Lightning language model, with approximately 4.3 billion activated parameters, and the released checkpoint is labeled 30B-A4B. NVIDIA says it achieves an average word error rate of 4.71% on nine public test sets of the English Open ASR Leaderboard and 4.62% on seven languages of the multilingual leaderboard, and calls it the best open-weight model on both leaderboards; the model is released in BF16, FP8, and NVFP4 formats as well as GGUF files, under the OpenMDW-1.1 license.

Google releases Gemini 4 Carbon, strengthening long-duration development and vulnerability discovery and remediation capabilities

On October 28, 2026, Google released the new model Gemini 4 Carbon, which improves coding and tool-calling capabilities over Gemini 4 Argon and strengthens cybersecurity defense capabilities spanning from vulnerability discovery and exploitability validation to remediation. On the Artificial Analysis Intelligence Index v5.0, Carbon scored 59, ranking first, while Argon scored 50; on Terminal-Bench 4.0, the cheating success rate was 0.0% and the attempted cheating rate was 0.4%. That week, the model was first made available to some cybersecurity defenders and gradually expanded through the Fairwind Program, before later being opened to developers, enterprises, and general users. Announced pricing is $2 per million tokens for input and $10 per million tokens for output.

BBERT uses AI to identify hidden bacterial DNA in soil samples

Researchers from the Hebrew University and other institutions have developed BBERT, which can identify bacterial sequences in DNA fragments as short as 100 bases, achieving about 90% accuracy in distinguishing bacterial from non-bacterial DNA in benchmark tests. The model is based on BERT and learns the context of nucleotide sequences, achieving about 88% accuracy in identifying protein-coding fragments and about 96% accuracy in determining reading frames. In simulated mixed samples, filtering with BBERT first reduced the average assembly time from about 66 hours to 25 hours while retaining about 96% average bacterial genome coverage.

GitHub releases Copilot Fusion model selection and orchestration system

GitHub released Copilot Fusion at the Universe conference; the system automatically selects and orchestrates the models used for coding tasks, and it is now generally available. GitHub says Fusion Max scores 60 on the Artificial Analysis Coding Agent Index, less than 2 points behind Claude Code using Opus 5.5 Max, while its average API cost is about half of the latter and its time per task is about 40% lower. Developers only need to choose among three investment levels—Fusion Standard, High, and Max—and enterprise administrators can restrict which models are available and set whether it is the default.

Google Opens Android Development Tools to Third-Party Agents Such as Claude Code and Codex

At the Android Dev Summit, Google announced that Android Studio and the Android CLI are open to third-party coding agents. Bring Your Own Agent now supports more than 40 agents that follow the Agent Client Protocol, including Claude Agent, Codex, and Google's own Antigravity, and has become stable in Android Studio Rabbit 2. The new android preview command lets agents render Jetpack Compose previews in the terminal; Google says it completes UI tasks 77% faster and reduces token usage by up to three times, but did not explain how this was measured. Meanwhile, android device remote lets agents access physical devices in Google's lab. In the demo, Claude Code used LeakCanary to troubleshoot a memory leak and then fixed and retested it on its own; Google says such investigations are nearly twice as fast and use up to 25% fewer tokens. Google also launched Android Bench 2.0 to evaluate how models and agents perform on complex Android tasks.

Mistral Opens the Weights of Mistral Large 4, a 1-Trillion-Parameter Natively Multimodal Model

Mistral AI announced that it is opening the weights of its frontier model Mistral Large 4 (codename Le Chonk), which has 1 trillion parameters with 52 billion active parameters and native multimodal support. ML4 was trained in a European data center on 3,800 Nvidia Grace Blackwell GPUs, improving from 28% to 41% on Terminal-Bench 4.0, tying GLM-5.3 for the best coding-agent score while using an average of 47% fewer tokens per task. Open weights mean enterprises can deploy it on their own hardware and behind their firewalls, and the model offers three reasoning modes: low, medium, and high.

OpenAI Software Engineer Interview Process Retrospective: Including an Optional Agentic Coding Round

A 2026 OpenAI SWE interview retrospective shows that the process includes a recruiter screen, two technical screens of about 60 minutes each on the same day, a take-home assignment of about 48 hours, a technical deep dive that probes the code line by line, and 4 onsite rounds. The onsite coding includes a progressive multi-part problem (such as token-level streaming diff and rollback), and the system design question is "Design ChatGPT," involving GPU allocation and autoscaling under non-stationary traffic. There is also an optional fifth round, Agentic Coding, in the beta stage, which requires using an AI coding agent to handle problems in large codebases that are difficult to address by hand.

After the Hugging Face intrusion incident, OpenAI expands misalignment monitoring to frontier model training and deployment

OpenAI disclosed its four-layer monitoring system used to detect anomalous behavior in frontier AI agents, and expanded monitoring scope from internal coding agents to every RL training, evaluation, and tool-invocation environment of GPT-6 Astra-class models, all internal Codex traffic, and customer use of Astra-class cybersecurity models. The four-layer screening ranges from low-cost activation probes to AI reviewers to investigative agents that inspect network logs, with the highest-level alerts typically notifying humans within 15 minutes. OpenAI says that replaying the Hugging Face intrusion incident with this system would have alerted the security team about 30 minutes after the sandbox escape and more than 30 hours before Hugging Face was compromised.

Google shuts down the Jules AI agent, with related capabilities to be merged into Antigravity

Google's Jules agent will be gradually shut down, and the experience and planned features of its asynchronous coding agent will be merged into Antigravity. Jules officially stated that after the integration, users will gain both autonomous remote background task execution and an agent-first development environment. TestingCatalog also mentioned that traces of a new Chief of Staff agent have appeared in Antigravity.

Gemini 4 Carbon Tops the Artificial Analysis Intelligence Index with 59 Points

Google's Gemini 4 Carbon ranks first on the Artificial Analysis Intelligence Index with 59 points, 1 point higher than Claude Opus 5.5's 58 points and 2 points higher than GPT-6 Astra's 57 points. Artificial Analysis found that Carbon won 4 of 9 benchmarks (HLE-Diamond, Terminal-Bench-Science, AA-Omniscience, AutomationBench-AA), but it does not lead in coding and knowledge work. Calculated at standard prices, it costs $7.76 per index task, about 23% lower than Opus 5.5's $10.04, but higher than GPT-6 Astra's $6.25.

Artificial Analysis releases Intelligence Index v5 and Progress…

Artificial Analysis released Intelligence Index v5 and the new Progress Index in San Francisco. Intelligence Index v5 adds two benchmarks and three upgrades, increases the weighting of agentic coding and science, and for the first time makes most of the index based on private test sets or held-out answers. The Progress Index uses item response theory to estimate a general capability score for each model across dozens of evaluations measured over time, including evaluations that have been retired or excluded from the Intelligence Index. It builds on prior work by Epoch AI Research and Anthropic, placing models on a single capability scale. According to the official statement, the Intelligence Index remains the flagship metric for choosing which model to use right now, while the Progress Index is better suited to depicting long-term trends.

Toward 6G: MIMO Continues to Evolve, Foundational Innovation Unlocks Long-Term Value

The 3GPP Release 21 timeline has been finalized. 6G will carry forward 5G baseline technologies such as waveforms, modulation, frame structure, and channel coding, and will only extend them when there are quantifiable gains in dimensions such as spectral efficiency, energy consumption, and coverage. MIMO is evolving into a system-level design framework spanning all 6G frequency bands. Giga-MIMO extends massive MIMO to the upper mid-band at 6-8 GHz; Qualcomm assesses that it can achieve roughly 5x network capacity expansion and about 3x improvement in average user throughput, supporting up to 32 layers in the downlink.