Anthropic releases Claude Sonnet 5, claiming its most agentic version yet approaches the Opus series
Anthropic's new Claude Sonnet 5 closes the gap to the pricier Opus model series
AI BriefingAnthropic has released Claude Sonnet 5, calling it the most agentic Sonnet version, able to plan autonomously and use browser and terminal tools. It surpasses Sonnet 4.6 across the reported benchmarks, scoring 63.2% on SWE-bench Pro, 80.4% on Terminal-Bench 2.1, and 1618 on the GDPval-AA v2 knowledge-work benchmark, slightly above Opus 4.8. It is available immediately on all plans, and developers can use it through Claude Code and the API, with a one-million-token context window. Through August 31, 2026, pricing is $2 per million input tokens and $10 per million output tokens, then returns to $3/$15. The model was not trained specifically for cybersecurity tasks, has real-time cyber defenses enabled by default, and shows less hallucination and sycophancy than its predecessor.
Also covered by 7 other sources
- Claude Sonnet 5 released as the Sonnet series' most capable agentic modelIT Home · Artificial Intelligence · July 1
- Claude Sonnet 5 releases a 145-page system card: Coding results and disclosures of multiple anomaliesRohan Paul · July 1
- What's new in Claude Sonnet 5Simon Willison's blog · July 1
- Anthropic releases Claude Sonnet 5: A mid-tier agentic model, with benchmarks and pricing explainedMarkTechPost · July 1
- Anthropic introduces Claude Sonnet 5, focusing on low-cost agentic capabilitiesTechCrunch: AI · July 1
- Claude Sonnet 5: The most agentic versionClaude · July 1
- Claude Sonnet 5 Released: 1M Context Window, Top-Tier Coding Tool PerformanceClaude Devs · July 1
Source: The Decoder: AI News · the-decoder.com