AI company Anthropic today officially launched Claude Sonnet 5, replacing Sonnet 4.6 released in February to become its most capable mid-range AI model. Positioned as the "mid-tier" option between the lightweight Haiku and the flagship Opus in the Claude product line, Sonnet 5 significantly boosts performance while maintaining a reasonable cost.
Anthropic calls Sonnet 5 its "most agentic Sonnet model to date." It can autonomously create plans, use tools like browsers and terminals, and execute tasks independently — achieving a level that just months ago required larger, more expensive models. Notably, Sonnet 5’s performance is close to Opus 4.8, but at a much lower price, narrowing the gap between mid-range and high-end models.
Starting today, Sonnet 5 becomes the default model for Claude’s free and Pro tier users. In terms of pricing, Anthropic is offering a limited-time deal: $2 per million input tokens and $10 per million output tokens, valid until August 31, 2026. This move further lowers the barrier to high-performance AI and is expected to attract more developers and enterprise users into practical applications.
Analysis: The launch of Sonnet 5 marks a key step for Anthropic on the "value-for-performance" track. As mid-tier model capabilities approach flagship levels, the trend of democratizing AI services will accelerate, giving developers more flexibility to balance performance and cost. Going forward, the blurring of model tiers may become an industry norm.
