The cost difference between AI coding models is actually insane.
GPT-6.1 Sol costs just $1.51 per coding task.
Meanwhile:
โข Claude Sonnet 5.5: $11.40 (7.5x more expensive)
โข Claude Opus 5.5: $12.61 (8.4x more expensive)
โข Gemini 4 Carbon (CLI): $6.84 (4.5x more expensive)
โข Gemini 4 Carbon (Antigravity): $7.13 (4.7x more expensive)
โข Grok 4.1: $15.35 (10.2x more expensive)
And seriously, who's even using Grok for coding? It's the most expensive model in this comparison, costing over 10x more than GPT-6.1
If your work mainly involves coding, I think Codex is probably the best option right now. I really don't see why you'd need to pay 7โ8x more just for coding tasks.