Theory: Sol is the new Terra and should be compared to Sonnet not Opus

A Reddit user theorizes that GPT-6 Sol, priced at $2 input/$10 output, should be compared to Anthropic's Sonnet 5.5, not Opus 5.5, given its new pricing structure which is half that of GPT 5.6 Sol and similar to Terra 5.6 and Sonnet 5. They suggest that OpenAI's model tiers (Luna, Sol, Astra) are clearer than Anthropic's, where the cheaper Opus sometimes outperforms Fable, making Sonnet a less distinct offering.

This theory uniquely compares GPT-6 Sol's new pricing, which is half that of GPT 5.6 Sol, to Anthropic's Sonnet 5.5, unlike other analyses focusing on Opus 5.5.