Anthropic released Claude Sonnet 5.5 on September 28, 2026, saying it generates output more than 30 percent faster and cuts per-task costs by up to 30 percent through more efficient token usage. It is the company's first major update in the Claude 5.5 family since the release of Opus 5.5.
Anthropic reported Sonnet 5.5 scores 70.6 percent on Terminal-Bench 4.0, compared to Sonnet 5's 10.3 percent. That compares with Opus 5.5's 66.4 percent on the same benchmark.
Sonnet 5.5 is built on Anthropic's internal architecture and targets well-defined everyday work like fixing bugs, writing docs, building presentations, and creating spreadsheets. Availability begins now, initially for developers on AWS, Google Cloud, and Azure.
"Sonnet 5.5 hits 70.6 percent on Terminal-Bench 4.0, which is a sharp jump from its predecessor," said Matthias Bastian, thedecoder's contributor. Anthropic claims the model's performance in coding benchmarks is significantly better.
The announcement follows the release of GPT-6 Luna and Sol by OpenAI. Anthropic's release comes amid a pricing battle with OpenAI's models, where performance differences between matched tiers are small enough that cost may end up being the deciding factor.
Anthropic did not say how independent testing will confirm these claims, and it raised the open question of whether the model's efficiency will hold under real-world conditions. The company said it will continue to refine the model and expand its capabilities.
Source: thedecoder
