ThinkingNews
Top Stories
7 outlets·15 reports·over 30h

Anthropic's Claude Opus 4.6 brings 1M token context and 'agent teams' to take on OpenAI's Codex

ThinkingNews Desk · how this was written

Anthropic has released Claude Opus 4.6, a model with a 1 million-token context window that adds “agent teams” capable of breaking larger tasks into smaller jobs and integrates directly into PowerPoint via a side panel. The upgrade is marketed for enterprise use, excelling in coding, tool use, financial analysis and security testing, and is offered at the same pricing as its predecessor.

Written from all 7 reports below, not from any single one.

How it was reported

  1. VentureBeat·
    Anthropic's Claude Opus 4.6 brings 1M token context and 'agent teams' to take on OpenAI's Codex

    Anthropic's Claude Opus 4.6 features a 1 million token context window and "agent teams" to enhance autonomous workflows. This upgrade outperforms competitors, including OpenAI's Codex, on key enterprise benchmarks. Opus 4.6 achieves the highest score on agentic coding evaluations.

  2. TechCrunch·
    Anthropic releases Opus 4.6 with new ‘agent teams’

    Anthropic's Opus 4.6 model is released with new 'agent teams'. This update broadens the model's appeal. Opus 4.6 enhances functionality.

  3. TechMeme·
    Anthropic releases Claude Opus 4.6, which it says can scrutinize company data, regulatory filings, and market info; Anthropic now has 300K+ business users (Rachel Metz/Bloomberg)

    Anthrropic releases Claude Opus 4.6, an AI model for financial research. It can analyze company data, regulatory filings, and market information. Anthropic now has over 300,000 business users.

  4. TechMeme·
    Anthropic says Claude Opus 4.6 supports a 1M context window, scored 90.2% on BigLaw Bench, the highest for any Claude model, and boosts agentic capabilities (David Gewirtz/ZDNET)

    Anthropic's Claude Opus 4.6 supports a 1M context window and scored 90.2% on BigLaw Bench. It boosts agentic capabilities for enterprise knowledge work. The model enables end-to-end autonomy with fewer rewrites.

  5. The Verge·
    Anthropic debuts new model with hopes to corner the market beyond coding

    Anthropic debuts Claude Opus 4.6, a model upgrade for complex tasks. It improves production-ready quality and excels in coding, tool use, and financial analysis. The new model is available with the same pricing as its predecessor.

  6. VentureBeat·
    OpenAI’s GPT-5.3-Codex drops as Anthropic upgrades Claude — AI coding wars heat up ahead of Super Bowl ads

    OpenAI's GPT-5.3-Codex is its most capable coding agent to date. The model achieves record-breaking benchmark scores, outpacing Anthropic's Claude. It aims to automate the entire software development lifecycle, transcending pure coding to general-purpose computing.

  7. TechMeme·
    Anthropic says Opus 4.6 adds "agent teams" that can split larger tasks into segmented jobs and integrates Claude directly into PowerPoint via a side panel (Lucas Ropek/TechCrunch)

    Anthropic released Opus 4.6, adding "agent teams" to split tasks into smaller jobs. The update integrates Claude into PowerPoint via a side panel. This enhances task management and accessibility.

  8. TechMeme·
    Anthropic says Opus 4.6 found 500+ previously unknown high-severity security flaws in open-source libraries with little to no prompting during its testing (Sam Sabin/Axios)

    Anthropic's Opus 4.6 found over 500 previously unknown high-severity security flaws in open-source libraries. The discovery was made with little to no prompting during testing. Opus 4.6 was tested in a sandboxed environment by Anthropic's frontier red team.

  9. TechMeme·
    Anthropic says it found Opus 4.6 "brings more focus to the most challenging parts of a task without being told to" and "thinks more deeply and more carefully" (Anthropic)

    Anthropic's Claude Opus 4.6 model brings more focus to challenging tasks. It thinks more deeply and carefully without being told to. The new model improves on its predecessor's coding skills.

  10. Hacker News·
    Anthropic's Claude Opus 4.6 uncovers 500 zero-day flaws in open-source code

    Anthropic's Claude Opus 4.6 discovered 500 zero-day flaws in open-source code. This highlights significant security vulnerabilities. Claude Opus 4.6's findings impact open-source security.

  11. TechCrunch·
    OpenAI launches new agentic coding model only minutes after Anthropic drops its own

    OpenAI launched a new agentic coding model to enhance Codex. This model accelerates Codex's capabilities, building on its earlier launch. The move follows Anthropic's release of a similar model.

  12. TechMeme·
    Anthropic details how it used 16 parallel Claude Opus 4.6 agents to build a Rust-based 100,000-line C compiler, incurring ~$20K in API costs over 2,000 sessions (Anthropic)

    Anthropic used 16 parallel Claude Opus 4.6 agents to build a Rust-based C compiler. The project involved 100,000 lines of code and incurred $20,000 in API costs over 2,000 sessions. This autonomous software development effort explored the future of agent teams in coding.

  13. TechMeme·
    Goldman Sachs says it has been working with Anthropic on AI agents to automate a growing number of roles at the bank, including client vetting and onboarding (Hugh Son/CNBC)

    Goldman Sachs is working with Anthropic to create AI agents. These agents will automate roles such as client vetting and onboarding. The partnership aims to increase automation at the bank.

  14. Wired·
    The Only Thing Standing Between Humanity and AI Apocalypse Is … Claude?

    Anthropic's philosopher believes Claude can learn wisdom to avoid AI disaster. The startup bets on Claude to prevent catastrophe as AI systems grow more powerful. Claude's development is crucial in avoiding an AI apocalypse.

  15. Ars Technica·
    Sixteen Claude AI agents working together created a new C compiler

    Anthropic's Claude AI agents created a C compiler. Sixteen agents built a 100,000-line Rust-based compiler over two weeks. The compiler can build a bootable Linux kernel on x86, ARM, and RISC-V architectures.

Related stories

Share: