Claude Sonnet 5.5 nearly matches Opus 5.5 at up to 30 percent less per task

Claude Sonnet 5.5 nearly matches Anthropic's flagship Opus 5.5 on benchmarks while costing up to 30 percent less per task, according to The Decoder. Anthropic pitches the new model as a significantly cheaper, faster work partner, TechCrunch reports.

What we know

Anthropic has released Claude Sonnet 5.5. Sonnet is the company's mid-tier line. Opus sits above it, with Opus 5.5 as the current top model.

The Decoder reports that Sonnet 5.5 comes close to Opus 5.5 on benchmarks. The gap between the two tiers has narrowed sharply.

The cost figure is measured per task. Running a job on Sonnet 5.5 instead of Opus 5.5 costs up to 30 percent less. That number is a ceiling. On some tasks, the savings will be smaller.

Sonnet 5.5 also improves on its own line. The German edition of The Decoder describes it as faster and cheaper than the previous Sonnet release.

Anthropic's own framing leans on speed and cost. TechCrunch quotes the company calling Sonnet 5.5 a significantly cheaper, faster work partner.

Why it matters

The old split was simple. Teams that needed the best results paid for the flagship. The mid-tier model was the compromise for work where speed and cost mattered more than the last bit of quality.

That math changes when the mid-tier model almost catches up on benchmarks. The question is no longer which model is stronger. It is which model is good enough for a given job, and which one costs less to run. Price becomes the real competition, and it starts inside Anthropic's own lineup.

For developers and companies that call models at high volume, per-task savings add up fast. Anyone running Opus today now has a concrete reason to rerun their workflows on Sonnet 5.5 and compare.

There is a second side for Anthropic. A mid-tier model this close to the flagship makes the Opus premium harder to justify. The company will need to show more clearly where the higher tier still earns its price.

Benchmarks also test standardized tasks. Whether the gap stays this small in daily work will only show in real use.

What's still unclear

The up-to-30-percent figure is a maximum, not an average, and typical savings remain unknown. It is also unclear whether they come from lower per-token prices, lower token use, or both. Which areas still favor Opus 5.5 by a clear margin has not been spelled out.