The story’s score and claims

Comments on No. 102 of 142 · GitHub · edition of Wednesday 18 November 2026

The model with the lower price per token cost me twice as much per task

Summary

A developer ran seven tasks through Qwen 4 Max and GPT-6 Sol, finding both passed all 93 checks but Qwen cost about $0.21 against $0.09 for GPT-6 Sol. The reason was Qwen's roughly five-times-higher output token count, mostly thinking tokens billed at the output rate, which outweighed its lower list…

Comments

Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules