GPT-6 Sol vs Claude Opus 5.5, plus Luna, at matched budgets: who scores higher per dollar, a hands-on coding test, speed, and how to route between them.
Key takeaways
- On the independent Artificial Analysis index, Claude Opus 5.5 scores 58 at its top setting, GPT-6 Sol 48 and GPT-6 Luna 37.
- Per token, Sol costs half as much as Opus 5.5, and Luna costs a fortieth as much.
- At matched cost, Sol wins up to a score of about 44. Above 48, only Opus 5.5 gets there.
- In one hands-on coding test at the high setting, Sol cost $0.26 and scored higher on quality than Opus 5.5 at $0.87.
- Anthropic's model at Sol's price, Claude Sonnet 5, scores far lower per dollar. A three-model routing plan beats any single choice.
📖 Read the full guide on Van Data Team → GPT-6 Sol vs Claude Opus 5.5, Plus Luna: Best Value per Task
