Anthropic launched Claude Sonnet 5 on June 30, making it the default model for Free and Pro plan users as of July 1. The model performs close to the flagship Opus 4.8 on many tasks but costs significantly less.

Introductory pricing runs through August 31: $2 per million input tokens and $10 per million output tokens. Standard pricing begins September 1: $3 per million input and $15 per million output.

The pricing is designed so that switching to Sonnet 5 is roughly cost-neutral despite the model’s power jump. Anthropic is betting that developers will keep the model after the intro period expires.

Performance vs. Price Tradeoff

Sonnet 5 delivers strong performance on reasoning, coding, math, and analysis tasks. It’s not Opus-level on every metric, but it’s close enough for most applications. For companies running high-volume inference, that’s a win.

The model uses an updated tokenizer that changes how text maps to tokens. Roughly 1.0–1.35× more tokens per input, depending on content type. That’s baked into the pricing math so cost-per-task doesn’t spike for users.

The Market Context

Anthropic is competing with OpenAI, Google, and others for developer mindshare. Price matters. Speed matters. Reliability matters. Sonnet 5 is Anthropic’s attempt to offer a sweet spot: capable enough for production use, cheap enough to run at scale.

The July 1 default shift means new Free and Pro users will use Sonnet 5 immediately. That’s millions of people suddenly getting a more capable model. Data collection begins instantly.

The Longer Game

By September 1, Anthropic will have data on how developers use Sonnet 5. If usage is strong and retention is high, the standard pricing increase becomes sustainable. If adoption lags, Anthropic might negotiate.

Sonnet 5 is Anthropic’s play for market share. The introductory pricing is bait. The real game is building developer loyalty that survives the price increase.