ANTHROPIC HAS LAUNCHED CLAUDE SONNET 5.5

Summary from the TLDR newsletter

Claude Sonnet 5.5 scores just 2 points behind Opus 5.5 (max) on the Artificial Analysis Intelligence Index. However, it uses around 60% more tokens to reach that performance. The model is priced at $2/$10 per million tokens of input/output, matching GPT-6 Sol. Sonnet 5.5 still lags on factual knowledge as it is a smaller class model than Opus 5.5, but it has a lower hallucination rate.