English text is being prepared.

English translation is being prepared.

English translation is being prepared.

View story:English translation is being prepared.

Mistral releases Large 4 weights on Hugging Face

On October 29, 2026, Mistral released the open weights of Mistral Large 4 (Le Chonk), a 1-trillion-parameter native multimodal model that continues reinforcement learning on the basis of the preview version, with improved coding and defensive cybersecurity capabilities. According to data published by Mistral AI, the model scores 19.7% on FrontierCode 2.0 and 48% on the Artificial Analysis Coding Agent Index, roughly on par with GLM-5.3, but uses an average of 47% fewer tokens per task; on Terminal-Bench 4.0 it improves by 13 percentage points to 41%. The model supports three reasoning modes, uses a modified MIT Mistral Ecosystem Open-Weight license, and is priced at $0.70 per million input tokens, $0.07 for cached input, and $2.10 for output via API. On the same day, Unsloth released a 1-bit GGUF quantized version of the model, stating that it can run locally on a Mac Studio, retains 91% top-1 accuracy after quantization, and provides a download link and usage guide.

GitHub announces that Copilot Fusion is officially available

On October 29, 2026, GitHub announced that Copilot Fusion is officially available. According to a report by Rohan Paul, the product can route coding tasks between Claude and GPT models; on Artificial Analysis's Coding Agent Index, its Max setting scored 60, compared with 62 for Claude Code, at a cost per task of $5.59 versus $11.42, with Fusion consuming 7.2M tokens per task, roughly half of Claude Code's usage. On the same day, Microsoft released multiple updates at GitHub Universe, stating that Copilot Fusion achieves frontier-level quality, is 40% faster, and reduces costs by up to 77%, and announced that a unified Copilot harness is entering Foundry as a managed service. There are currently no subsequent developments.

Shao Meng recounts a review of the 2026 OpenAI SWE interview process

Shao Meng published a personal retrospective of the 2026 OpenAI software engineer (SWE) interview process. The retrospective shows that the process includes recruiter screening, 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 a 4-round onsite stage. The onsite coding is a progressive multi-part problem involving token-level streaming diff and rollback; the system design question is "Design ChatGPT," involving GPU allocation and autoscaling under non-stationary traffic. There is also an optional fifth round in beta, Agentic Coding, which requires using an AI coding agent to handle problems in large codebases that are difficult to address by hand. At present, this retrospective is Shao Meng's unilateral account, and no other sources have cross-corroborated it.

OpenRouter launches the Particle-1 model

OpenRouter has launched the Particle-1 model. According to Particle, the model is post-trained on DeepSeek V4.1 Flash and uses 28–52% fewer output tokens than DeepSeek V4.1 Flash on four coding benchmarks, an average of about 45% fewer tokens, with accuracy close to the base model. Its pricing is $0.25 per million input tokens and $1 per million output tokens, with a context window of 1.05M. There are no further developments at present.

Larsen Cundric says users are shifting more toward desktop and mobile

On October 30, 2026, Larsen Cundric made an observation, stating that compared with six months earlier there had been a noticeable change: people are increasingly interacting with coding agents through desktop and mobile, and he asked in return whether developers are still staying in the terminal. Previously, he had launched a poll asking developers whether they actually talk to coding agents in the terminal, IDE, desktop app, or mobile app. At present, this claim comes only from content he himself published; there are no independently developed facts from other sources.

Serro launches Agentic TPM

Serro has released "Agentic TPM," designed to keep coding agents in sync when the decisions they rely on change. The product connects to Slack, Linear, GitHub, Jira, Zoom, Notion, email, and code repositories, maintains a versioned record of each project's decisions, owners, risks, and history, and treats agent failures as coordination problems. The news was reported by Rohan Paul on October 30, 2026, and there have been no further developments so far.

Nathan Lambert reviews the Artificial Analysis Intelligence Index v5

On October 28, 2026, Nathan Lambert published a commentary on the newly released Intelligence Index v5 from Artificial Analysis. The index weights knowledge work, coding, science, and general categories at 25% each, and provides new model rankings. Lambert noted that eight closed-source models in the rankings lead the best open-source model, Kimi K3.1 Pro, arguing that the new index leans more toward measuring where frontier models are heading and may underestimate the role of open-source models in the economy. On the same day, he said he wanted to defend both the AAv5 index and open-source models, and argued that benchmarks are messier, more subjective, and trickier than most people think. The article also discusses the pros and cons of shifting to private benchmarks, and how the business model of benchmark organizations charging model vendors to verify results affects credibility.

Coforge launches Autonomous Software Factory

Coforge launches 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. There are currently no subsequent developments.

Hebrew University team publishes BBERT bacterial DNA identification research

Researchers from the Hebrew University and other institutions have published a paper titled "Fast and accurate taxonomic domain assignment of short metagenomic reads using BBERT," proposing the BBERT model. This model can identify bacterial sequences in DNA fragments as short as 100 bases; in benchmark tests, its accuracy in distinguishing bacterial from non-bacterial DNA is about 90%, its accuracy in identifying protein-coding fragments is about 88%, and its accuracy in determining the reading frame is about 96%. 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. The research has currently been published as a paper, and no follow-up progress has been reported.

Google launches the BYOA feature in Android Studio Rabbit 2

At the Android Dev Summit, Google launched the Bring Your Own Agent (BYOA) feature for Android Studio Rabbit 2. Through the open Agent Client Protocol, developers can run more than 40 coding agents simultaneously within the IDE, including Claude Agent, Codex, and Google Antigravity, with direct access to the project structure, emulator, and Logcat. Google says that in internal testing, agents completed UI tasks 77% faster after rendering Compose previews in the terminal, with token usage reduced to as little as one-third; in an A/B test using Claude Opus 5.5, the built-in LeakCanary CLI found and fixed memory leaks nearly twice as fast, with about 25% fewer tokens. There are currently no further developments.

Mistral calls Le Chonk the world's best open-weight model

On October 29, 2026, Mistral released the official version of Le Chonk, a trillion-parameter open-weight model, calling it the world's best open-weight model for automated coding and cybersecurity, available for free. Pierre Stock, Mistral's Vice President of Science, said the model was trained on increasingly difficult cybersecurity and coding problems, enabling it to match or surpass models from Chinese labs such as DeepSeek and Z.ai. Stock also emphasized that the West needs its own advanced open-weight models, because closed-source models can be shut down remotely, whereas open weights cannot have access cut off once released. Currently, this event has only been reported by Wired AI, with no further developments yet.

Mastra launches Span Queries

On October 28, 2026, Mastra launched Span Queries, which can filter and group individual spans across all traces by conditions such as time range, type, name, status, error, and duration, to help locate common failures in coding agents. This feature is available in Mastra platform, Studio, CLI, Client SDK, and observability endpoints. Each query can combine up to 100 conditions, return up to 1,000 spans per page, group by up to 2 dimensions, and cover a time range of up to 365 days; the CLI requires mastra@1.36.0 or later, and the SDK requires @mastra/client-js@1.57.0 or later. Self-hosted projects can also use it when storage.observability uses DuckDB, PostgreSQL, or ClickHouse. There are currently no further developments.

Amazon Updates Software Engineer Working Guidelines for the AI Era

On October 28, 2026, Amazon updated its working guidelines for software engineers for the AI era. According to a TechSpot report, the guidelines require teams to measure their value by the software delivered with the help of AI coding tools rather than by token consumption, and warn against treating AI adoption rates, token usage, code changes, and code reviews as performance metrics; the company will still look at usage data, but it must be combined with other metrics. The report said the guidelines target tokenmaxxing behavior—deliberately inflating AI consumption to appear efficient on internal dashboards—and allow engineers to conclude that a given task is not suited to the available tools. 36Kr·AI reported on October 30 that this guidance document was disclosed by Vice President Terk Terkowitz and introduces the concept of the "Renaissance developer," requiring technical leaders to personally use AI in production work and to explain the scenarios in which it is effective and ineffective.

Rohan Paul suggests that tool selection is becoming one of the most critical decisions for Agents

Rohan Paul puts forward the judgment that tool selection is becoming one of the most important decisions for Agents, often more important than which model writes the final answer. This judgment appeared in his October 28, 2026 article about AI/ML API launching Jev Router: the router places more than 10,000 tools and models behind a single MCP server, with Typesafe's Jev deciding who handles each task; the agent only needs to connect once, does not need to hardcode a provider for each step, and does not need to carry a tool catalog in the prompt. The company claims it can complete 28 times more tasks per dollar than Claude and answer 4 times more questions per minute. There are currently no follow-up developments.

Google releases Gemini 4 Argon and demonstrates complex task capabilities

On October 27, 2026, Google released its latest model, Gemini 4 Argon, and demonstrated its ability to handle complex real-world tasks. Google says the model can reason, code, and coordinate tools, and the company gave 8 examples: first, writing 5,000 lines of mathematical code from scratch with zero source material, rendering a 3D sunlit library; second, processing more than 80,000 raw files from the NASA Mars mission in Antigravity and generating executive briefing slides. There is currently no information on further developments.

Mistral calls Le Chonk the world's best open-weight model

On October 29, 2026, Mistral released the official version of Le Chonk, a trillion-parameter open-weight model, calling it the world's best open-weight model for automated coding and cybersecurity, available for free. Pierre Stock, Mistral's Vice President of Science, said the model was trained on increasingly difficult cybersecurity and coding problems, enabling it to match or surpass models from Chinese labs such as DeepSeek and Z.ai. Stock also emphasized that the West needs its own advanced open-weight models, because closed-source models can be shut down remotely, whereas open weights cannot have access cut off once released. Currently, this event has only been reported by Wired AI, with no further developments yet.

View story:Mistral calls Le Chonk the world's best open-weight model

AISI and SecureBio Build ReproBAIT Evaluation

The UK AI Security Institute's Chem-Bio team, in collaboration with SecureBio, built the ReproBAIT (Reproducing Biological AI Tools) evaluation to test the ability of AI coding agents to reproduce biological AI tools given data and limited public information. The results showed that simpler tools could be reproduced within 4 hours and under a $20 compute budget; ProteinMPNN reached 96.5% reproduction fidelity with Claude Opus 5 when the full paper was provided; and for the most complex RFdiffusion3, no agent succeeded in reproducing it within the time limit. The team therefore noted that releasing only code or restricting paper information does not equal safe release, and that controlling access to the underlying biological data may be more effective than controlling code.

Thariq announces departure from Anthropic

Thariq, a core developer of Claude Code, announced that he is leaving Anthropic. He said he worked at Anthropic for over a year, was involved with Claude Code for more than a year, and witnessed Claude Code scale far beyond what he had imagined. Claude Code is an autonomous coding agent that resides in the CLI and can branch on its own, read codebases, locate bugs, run tests, and submit git commits. Thariq said he still believes in the company's mission and "race to the top," and that he will take a short break before starting something new, and recommended following @addyosmani, @lydiahallie, @_catwu, @ClaudeDevs, and @bcherny. There are currently no further developments.