Anthropic has released Claude Sonnet 5.5, the second model in its Claude 5.5 family, bringing speed and performance upgrades for everyday tasks, software development, and complex document creation.
The new model runs more than 30 percent faster than its predecessor, Claude Sonnet 5, while cutting overall task expenses by up to 30 percent.
Sonnet 5.5 serves as a lower-cost complement to the high-judgment Claude Opus 5.5 model, with the high-volume Claude Haiku 5.5 scheduled for release in the coming weeks.
Benchmark Gains Across Coding and Knowledge Tasks
Benchmark testing shows substantial performance gains across multiple technical domains.
On the Terminal-Bench 4.0 agentic coding evaluation, Sonnet 5.5 scored 70.6 percent, up from 10.3 percent posted by Sonnet 5.
The model also reached within two points of Opus 5.5 on GDPval-AA, an evaluation measuring real-world workplace tasks across 44 occupations and nine industries.
In programming benchmarks, it scored 10 points higher than Sonnet 5 on FrontierCode at High effort settings while reducing task costs by approximately fifteen-fold.
It also outperformed previous iterations on long-horizon knowledge work and image processing, becoming the first Sonnet model to navigate Pokémon Red using only screenshots.
Early corporate evaluators noted enhanced operational efficiencies during technical testing, including improved codebase comprehension and expanded tool-call batching that minimized total execution steps.
"In Epic’s early testing, Claude Sonnet 5.5 cleared the same quality bar you’d expect from a higher-tier model, holding up on a system design audit and a data flow review," an evaluator said.
"The new model managed tens of thousands of lines of code for gameplay system architecture, kept responses snappy, handled multi-hour tasks, and delivered with less prescriptive prompting."
API pricing remains identical to Sonnet 5 at $2 per million input tokens, $10 per million output tokens, and $0.20 per million cache read tokens, though reduced overall token usage lowers per-task expenses.
Anthropic also equipped Sonnet 5.5 with cybersecurity safeguards comparable to Opus 5, marking the first Sonnet variant to deploy specialized safeguards targeting high-risk requests.
