Anthropic AI Updates: July 1, 2026

1. Anthropic Introduces Claude Sonnet 5

Anthropic. Anthropic released Claude Sonnet 5, a model that brings agentic performance close to Opus 4.8 at lower cost, with substantial gains over Sonnet 4.6 in reasoning, tool use, coding, and knowledge work. The model is priced at an introductory rate of $2 per million input tokens and $10 per million output tokens through August 31, 2026, and Anthropic positions it for complex multi-step software engineering and automation workflows. Sonnet 5 also becomes the default model in Claude Code. Source

2. Claude Science Launches as an AI Workbench for Researchers

Anthropic. Anthropic launched Claude Science, a customizable workbench that consolidates fragmented research tooling into a single environment with over 60 curated skills and connectors pre-configured for domains including genomics, single-cell analysis, proteomics, structural biology, and cheminformatics. The platform produces auditable artifacts and provides flexible access to computing resources, and it is available in beta to Claude Pro, Max, Team, and Enterprise users. Anthropic is offering grant funding of up to $30,000 in credits for qualifying research projects. Source

3. Anthropic Redeploys Fable 5 After Export Controls Are Lifted

Anthropic. Anthropic announced that export controls on Claude Fable 5 and Mythos 5 have been lifted, enabling global redeployment beginning July 1, 2026. The models were briefly restricted after the US government imposed controls on June 12 following the discovery of a jailbreak technique by Amazon researchers, and Anthropic has since deployed improved safety classifiers to address the vulnerability. The company said it is partnering with major technology firms to develop an industry framework for assessing the severity of AI jailbreaks. Source