Anthropic has released Claude Sonnet 5, a new midrange AI model that the company says can handle complex jobs more independently than earlier Sonnet versions. According to SiliconAngle, the launch happened on June 30, 2026, positioning the model as an upgrade over its predecessor in coding and safety.
The pitch centers on autonomy. Anthropic says Sonnet 5 can create plans, switch between different tools when needed, and keep working through a task on its own. Fonearena reports the model is designed for "agentic" workflows, using tools such as browsers and terminals to operate with less hand-holding.
Performance is the headline claim. Multiple outlets, including IT Pro, note that Anthropic is touting results "close to that of Opus 4.8, but at lower prices" — its more expensive flagship tier. Startup Fortune framed it as bringing near-Opus performance to developers at a fraction of the cost, while Tech My Money described the release as narrowing the gap between the cheaper Sonnet line and the premium Opus line.
Availability is broad. According to The Hans India, Anthropic has made Sonnet 5 the default model for users on both Free and Pro plans, pairing wider access with lower pricing.
The move also lands in a competitive context. VOI.id characterized the launch as warming up an "AI agent price war," and Quartz reported that Anthropic is making AI agents cheaper to run. An MSN item noted the release reflects a broader shift across the AI industry toward models that can complete work autonomously.
Why it matters: by pushing near-flagship capability into a cheaper, default-access model, Anthropic is lowering the cost of running autonomous AI agents — a step that could make hands-off AI tools far more widely usable, and intensify price competition among AI providers.