On June 30, 2026, Anthropic released Claude Sonnet 5, calling it the most agentic Sonnet model yet. The company says the model shows substantial improvements over Sonnet 4.6 in reasoning, tool use, coding, and knowledge work, and that it can make plans, use tools like browsers and terminals, and run autonomously at levels that previously required larger models.
Anthropic positions Sonnet 5 as approaching the performance of its frontier Opus 4.8 model at a much lower price. The announcement cites gains on agentic search (BrowseComp) and computer use (OSWorld-Verified) evaluations, with substantially improved cost efficiency at medium effort. Introductory API pricing is 2 dollars per million input tokens and 10 dollars per million output tokens through August 31, 2026, after which standard pricing of 3 dollars in and 15 dollars out applies.
Sonnet 5 shipped immediately across all plans: it became the default model for Free and Pro users and is available to Max, Team, and Enterprise customers, in Claude Code, and on the Claude Platform via the model ID claude-sonnet-5. On safety, Anthropic reports a lower overall rate of undesirable behaviors than Sonnet 4.6, better refusal of malicious requests, and deliberately weaker cybersecurity exploit-development capability than Opus 4.8, with cyber safeguards enabled by default.
For businesses, the release continues a clear industry pattern: capabilities that were frontier-priced six months earlier keep migrating into cheaper midsize models, which is where most production agent workloads actually run.