What happened
Anthropic released Claude Opus 4.6 with a native 1 million token context window, making it the first frontier model to ship with megascale context as a default capability rather than an experimental feature. The model maintains strong performance on needle-in-a-haystack benchmarks across the full context length and introduces improved agentic tool use, with Claude Code receiving same-day integration as its default model.
Why it matters
A 1M context window changes what agents can do in a single pass. Entire codebases, full document repositories, and multi-hour conversation histories fit in context without chunking or RAG workarounds. For Claude Code users specifically, this means the agent can hold an entire project in memory while making cross-cutting changes — no more losing track of distant files. The practical ceiling for single-shot agentic tasks just moved significantly upward.
Who should pay attention
AI engineers building agent systems, Claude Code power users working on large codebases, enterprise teams evaluating LLMs for document-heavy workflows, and anyone building RAG systems who should now re-evaluate whether they still need retrieval for their use case.