Anthropic Web Configuration Suggests Advancements in Claude Sonnet 4 and Opus 4 Technologies

Anthropic is reportedly advancing its efforts with new AI models labeled Claude Sonnet 4 and Opus 4, following recent clues extracted from the company’s web configuration files. According to these configurations, references to these models hint at their status as potentially the organization’s most sophisticated iterations yet.

The insights have emerged from the adjustments made to Anthropic’s internal files, which explicitly mention “Claude 4,” “Claude Sonnet 4,” and “Claude Opus 4.” This suggests that these models are part of the Claude 4 series and may represent Anthropic’s pinnacle of AI development.

The internal documentation implies that Claude Sonnet 4 and Opus 4 are currently going through testing or evaluation phases, indicated by disclaimers such as “Not intended for production use” and “strict rate limits.” These warnings suggest that the models remain under active development or are available in limited early-access formats.

Further details from the configurations allude to advanced features, including “showrawthinking,” which suggest enhanced reasoning capabilities inherent in these new models.

In related developments, Anthropic is scheduled to host a ‘Code with Claude’ event on May 22, though it remains uncertain if this occasion will coincide with an announcement regarding the new models.