Anthropic launches Claude Sonnet 5, an agentic model that costs less than flagship Opus 4.8

On June 30, 2026, Anthropic announced Claude Sonnet 5, the most agentic mid-tier model in the company's lineup. On many tasks, it comes close to the quality of the flagship Claude Opus 4.8, but it costs less than half as much. The company said Sonnet 5 is available today on all plans.

What Claude Sonnet 5 can do

According to Anthropic, the model was designed first and foremost to act, not just answer questions. The focus is on agentic scenarios: running multistep tasks, using tools, working with code and files, and carrying workflows through to completion without a human stepping in at every stage. That sets Sonnet 5 apart from earlier generations of mid-tier models, which were optimized mainly for the quality of a single response rather than for reliably executing long chains of actions.

• Launch date: June 30, 2026
• Claude Sonnet 5 is available immediately on all Anthropic plans, including the free tier
• On many tasks, the model comes close to the quality of the flagship Claude Opus 4.8
• Sonnet 5 costs less than half the price of Opus 4.8

How much Sonnet 5 costs compared with Opus 4.8

The new model's main selling point is its price-to-performance ratio. Anthropic positions Sonnet 5 as a model that handles most real-world tasks at nearly flagship level while costing companies that run AI agents at scale half as much. For businesses whose token bills grow with the number of agents they run and tasks they process, those economics directly shape the decision to move from experiments to production. Agentic workflows typically call the model many times per task — to plan a step, invoke a tool, check the result — and the per-token price gets multiplied by every one of those iterations.

Why agentic models are becoming a product category of their own

Sonnet 5 extends a trend that every major AI lab is setting in 2026: instead of one general-purpose model, companies ship lineups tiered by price, speed, and quality, with a specific focus on autonomous task execution — from writing and editing code to operating a browser and companies' internal systems. In this segment, Anthropic competes with agentic products from OpenAI and Google, and the price per agentic token is becoming as important a competitive factor as a model's raw benchmark quality.

What it means

Cheaper mid-tier agentic models lower the barrier to running AI agents in production: companies no longer need to pay for a flagship model where Sonnet 5's quality is good enough. That speeds up the race among labs for the best price-to-performance ratio in agentic AI.

FAQ

How is Claude Sonnet 5 different from Claude Opus 4.8?

Sonnet 5 is a mid-tier model that comes close to the quality of the flagship Opus 4.8 on many tasks but costs less than half as much; Opus 4.8 remains Anthropic's more powerful flagship model.

When did Claude Sonnet 5 become available?

Anthropic announced the launch on June 30, 2026, and opened access to the model on all of the company's plans the same day.