Anthropic Releases Claude Opus 4.6 and Sonnet 4.6 With Leading Coding Benchmarks

Summary

Anthropic released Claude Opus 4.6 on February 5, followed by Sonnet 4.6 on February 17. Opus 4.6 achieved 80.8% on SWE-bench (a new high), introduced a 1M context window (beta) — a first for Opus-class models — and 128K output tokens. New features include agent teams in Claude Code (research preview), context compaction for automatic summarization during long-running agentic tasks, and four-level effort controls. Sonnet 4.6, positioned as "Opus-level performance at Sonnet prices," was preferre