Claude Opus 4.6 Launches with Agent Teams, 1M Context Window, and 128K Output

Anthropic's Most Powerful Model Yet

Anthropic released Claude Opus 4.6 on February 5, 2026, marking a significant leap in AI capability. The flagship model introduces several groundbreaking features that push the boundaries of what AI coding agents can achieve.

Key Features

Agent Teams (Research Preview): Multiple Claude Code instances can now coordinate on complex tasks through a tmux-based orchestrator pattern. One session acts as team lead, assigning tasks and synthesizing results while teammates work independently. In one experiment, 16 parallel Claude agents wrote a 100,000-line C compiler in Rust in just two weeks, achieving a 99% pass rate on the GCC test suite.

1 Million Token Context Window: Opus 4.6 scores 76% on MRCR v2, a needle-in-a-haystack benchmark, compared to just 18.5% for Sonnet 4.5 — enabling full codebase analysis in a single prompt.

128K Output Tokens: Doubled output capacity allows Claude to generate substantially longer responses for complex tasks.

Adaptive Thinking: Claude now dynamically determines when and how much to use extended thinking based on request complexity, with four effort levels (low, medium, high, max).

Context Compaction (Beta): Automatic server-side conversation summarization enables effectively infinite conversations without hitting context limits.

Benchmark Performance

On GDPval-AA — measuring economically valuable knowledge work — Opus 4.6 outperforms OpenAI's GPT-5.2 by approximately 144 ELO points. The model also integrates natively with PowerPoint and Excel.

Source: Anthropic - Introducing Claude Opus 4.6 | TechCrunch | VentureBeat | MarkTechPost