GPT-6 Luna (Low) vs GPT-6 Sol (High)
Which is better for the money: GPT-6 Luna (Low) by OpenAI or GPT-6 Sol (High) by OpenAI? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-26.
Trade-off
Verdict
Depends on your budget
GPT-6 Sol (High) scores 20.9 points higher (42.4 vs 21.5) but costs 57× as much to run the benchmark ($605 vs $10.6; premium vs budget cost classes). They are not alternatives at one budget: pick GPT-6 Sol (High) when you need intelligence above 21.5, and GPT-6 Luna (Low) when 21.5 is enough.
| Metric | GPT-6 Luna (Low) | GPT-6 Sol (High) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 21.5 | 42.4 | +97.2% GPT-6 Sol (High) is 97.2% more intelligent |
| Total AA cost (lower is better) | $10.63 | $604.52 | +5586.9% GPT-6 Sol (High) is 57× more expensive |
| End-to-end time (lower is better) | 7.32s | 21.24s | +190.2% GPT-6 Sol (High) is 190.2% slower |
| Price per 1M tokens (in / out) | $0.10 / $0.50 | $2.00 / $10.0 | |
| Input modalities | Text, Image | Text, Image |
GPT-6 Sol (High) vs GPT-6 Luna (Low): 97.2% more intelligence, 5586.9% more cost to run the AA benchmark, and 190.2% slower end-to-end.
Today's Pareto frontier
Against the frontier
- GPT-6 Luna (Low) is on today's frontier: nothing is both smarter and cheaper.
- GPT-6 Sol (High) is off the frontier: at the same or lower cost, GPT-6.1 Sol (High) scores 50.20 for $521.
GPT-6 Luna (Low)GPT-6 Sol (High)Today's Pareto frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/gpt-6-luna-low/gpt-6-sol-high · Markdown: /api/v1/compare/gpt-6-luna-low/gpt-6-sol-high?format=md · llms.txt · agent skill