Anthropic released Claude Sonnet 5 on 30 June, making it the default model for every Free and Pro Claude user from 1 July — the company's biggest mass-market model launch of the year and a signal that the AI competition is shifting decisively from chat to agentic capabilities.
Sonnet 5 is described as the most agentic Sonnet model ever built, with performance close to the flagship Opus 4.8 on coding, debugging and multi-step work — tasks that define the emerging agent-first paradigm. The model excels at sustained, multi-turn problem-solving where earlier Sonnet versions lost coherence or required manual intervention.
The pricing undercuts the market aggressively. Introductory rates are set at $2 per million input tokens and $10 per million output tokens through 31 August 2026, after which pricing moves to $3 per million input and $15 per million output. At launch pricing, Sonnet 5 is roughly half the cost of previous Sonnet models while delivering substantially better agentic performance.
The model is available across all plans — Free, Pro, Max, Team, and Enterprise — removing the capability gap that previously existed between free and paid tiers. This democratisation strategy puts frontier-adjacent AI capabilities in the hands of every Claude user, regardless of subscription level.
The launch timing is strategic. It arrives just days after Anthropic overtook OpenAI in business subscriptions and with ChatGPT's market share falling below 50 per cent for the first time. Making a near-Opus-quality model the default for free users is a distribution play — converting casual users into committed Claude users before competitors can respond.
The shift from chat to agents has broader implications. Chat models optimise for single-turn responses; agentic models optimise for sustained, multi-step task completion. Sonnet 5's design reflects Anthropic's bet that the primary value of AI is not answering questions but completing work — a distinction that shapes everything from pricing to product design.
For context engineers, Sonnet 5's launch as the default free model means that the baseline capability available to every developer has taken a significant step forward. Multi-step coding tasks, debugging sessions, and agentic workflows that previously required Opus-tier models are now accessible at Sonnet pricing — a shift that lowers the cost of AI-assisted development for the entire ecosystem.
