Anthropic released Claude Sonnet 5.5 on Monday, an upgrade to Sonnet 5 from June, at the unchanged price of $2 per million input tokens and $10 per million output tokens, half the cost of Opus 5.5. The model scored 70.6% on Terminal-Bench 4.0 versus 66.4% for Opus 5.5 according to Anthropic, and 63.6% versus 59.6% according to independent lab Artificial Analysis. Sonnet 5.5 runs over 30% faster than its predecessor and uses nearly a third fewer tokens per task, making it cheaper to run than Sonnet 5. However, at maximum effort level, it generates approximately 193,000 tokens per test task, the highest volume ever measured by Artificial Analysis, costing around $7.60 per task.
Source: Read the original article

