Claude Sonnet 5 (Vertex AI (Anthropic))

A Claude timeline page lists Claude Sonnet 5 with a release date of June 30, 2026, describing it as Anthropic's Sonnet model for agentic coding, tool use, reasoning, knowledge work, and lower-cost agent workflows. The current Anthropic API lineup includes Claude Fable 5.1, Claude Opus 5, Claude Sonnet 5, and Claude Hai For context, the timeline places Sonnet 5 between Claude Opus 5 (July 24, 2026) and Claude Fable 5 (June 9, 2026) in the Claude model lineup. Claude Fable 5.1, released September 1, 2026, is described as Anthropic's most capable generally available model with a 1M-token context window and 128K maximum output. Claude Op

Anthropic officially released Claude Sonnet 5 on June 30, 2026, positioning it as "the most agentic Sonnet model to date." According to the report, Sonnet 5 can plan tasks, use tools such as browsers and terminals, and operate autonomously at a level that previously required larger, more expensive models. It shows sign The article details cost-performance trade-offs across effort levels, showing Sonnet 5 with clear gains over Sonnet 4.6 in the agent search evaluation BrowseComp and the computer use evaluation OSWorld-Verified. At medium effort levels, Sonnet 5 significantly improves cost-efficiency; at higher effort levels, it can ma

Anthropic introduced Claude Sonnet 5 on June 30, 2026, positioning it as the most agentic Sonnet model yet, capable of planning, using tools like browsers and terminals, and running autonomously. It is available across all Claude plans and priced at $2 per million input tokens and $10 per million output tokens. The model narrows the gap with Opus 4.8 at lower prices, showing substantial improvements over Sonnet 4.6 in reasoning, tool use, coding, and knowledge work. Safety assessments indicate a lower rate of undesirable behaviors than Sonnet 4.6 and reduced cybersecurity capabilities compared to current Opus models.

An aggregated specs and benchmarks page for Claude Sonnet 5, updated September 21, 2026, lists the model as Anthropic's Sonnet-class LLM for reasoning, tool use, coding, computer use, and knowledge work. Key specifications include a 1M-token context window, an Anthropic API input price of $2, an overall LLMBoard Score Detailed per-benchmark results include SWE-Bench Verified at 85.20% (rank 5 of 116, 96.52 percentile), OSWorld-Verified at 81.20% (rank 6 of 24, 78.26 percentile), Terminal-Bench 2.0 at 80.40% (rank 3 of 53, 96.15 percentile), and USAMO 2026 at 33.39 points (rank 2 of 3). These benchmark scores provide a quantitative v