Claude Sonnet 5

Caylent's launch analysis frames Sonnet 5 as raising the Sonnet baseline while nearly catching up to Opus 4.8 capability at the Sonnet price, making it a strong upgrade path for Sonnet 4.6 workloads that need stronger coding, tool use, and agentic follow-through. The piece restates the three launch behavior changes doc The article recommends task-level enterprise validation against real codebases, tools, documents, latency, cost, and safety requirements rather than treating the launch as a blanket upgrade, since some workloads will benefit more than others. Sonnet 5 is described as available across Claude plans, Claude Code, Claude P

Complete list of Anthropic latest updates for July 2026: get every product news, release note, and changelog from Anthropic summarized in one timeline.

Release notes ; July 1, 2026. Access to Claude Fable 5 and Claude Mythos 5 restored ; June 30, 2026. Claude Sonnet 5 launch ; June 25, 2026. Enable ...

Anthropic released Claude Sonnet 5 on 30 June 2026, framing it as the most agentic Sonnet model to date, with introductory API pricing of $2 per million input tokens and $10 per million output tokens through 31 August 2026, stepping up to $3/$15 standard pricing after that date — identical on a per-token basis to what The cost gap is attributed by the analysis to two changes bundled into the release: materially higher token consumption from the new tokenizer (which the official Anthropic docs quantify at roughly 30% more tokens per identical input), and longer agentic reasoning traces from the now-default adaptive thinking mode. On

Dojo Labs' analysis (updated September 13, 2026) corroborates Anthropic's June 30, 2026 launch positioning of Sonnet 5 as the most agentic Sonnet model to date, designed to narrow the gap with Opus 4.8 at lower cost. It reports Sonnet 5 launch pricing of $2/M input and $10/M output through August 31, 2026, then $3/M in The post tabulates benchmark results: SWE-Bench Pro 63.2% (up from Sonnet 4.6's 58.1%), Terminal-Bench 2.1 80.4% (up from 67.0%, the largest single gain in the comparison), OSWorld-Verified 81.2% (up from 78.5%), and Humanity's Last Exam with tools at 57.4%, close to Opus 4.8's 57.9%. It also flags a tokenizer change w

Anthropic launched Claude Sonnet 5 on 2026-06-30 as its most agentic Sonnet model yet, shipping via the Claude Platform API under the model id claude-sonnet-5 and becoming the default for Free and Pro plans while remaining available to Max, Team, and Enterprise tiers in Claude Code and the API. The launch introduces a Sonnet 5 intro pricing is $2 per million input tokens and $10 per million output tokens through 2026-08-31, stepping to $3 / $15 per MTok from 2026-09-01, compared with Opus 4.8 at $5 / $25 per MTok. On BrowseComp, Sonnet 5 reaches roughly 82–85% pass rate at high/xhigh effort, overlapping Opus 4.8; on OSWorld-Verified

Anthropic released Claude Sonnet 5 on June 30, 2026, describing it as the most agentic Sonnet model to date and positioning it between Sonnet 4.6 and Opus 4.8 on capability at Sonnet pricing. The API model ID is `claude-sonnet-5`, and the model is the default for Free and Pro plans in the Claude app, with availability For developers, launch API pricing is reported at $2 per million input tokens and $10 per million output tokens through August 31, 2026, after which standard pricing is $3 per million input tokens and $15 per million output tokens. As a third-party paraphrase of Anthropic's launch post and system card, these pricing an

Sonnet5 is Anthropic's most capable Sonnet model to date in terms of agent performance and has undergone new safety optimizations. It is now the default model on the Claude platform, available to all users. Sonnet 5 delivers performance close to that of the flagship Opus 4.8 across a wide range of agent tasks and, duri