Claude Opus 5 (Medium) vs GPT-5.6 Luna (High)
Which is better for the money: Claude Opus 5 (Medium) by Anthropic or GPT-5.6 Luna (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
Claude Opus 5 (Medium) scores 12.7 points higher (44.8 vs 32.1) but costs 25× as much to run the benchmark ($2732 vs $108; flagship vs mid-range cost classes). They are not alternatives at one budget: pick Claude Opus 5 (Medium) when you need intelligence above 32.1, and GPT-5.6 Luna (High) when 32.1 is enough.
| Metric | Claude Opus 5 (Medium) | GPT-5.6 Luna (High) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 44.8 | 32.1 | −28.3% GPT-5.6 Luna (High) is 28.3% less intelligent |
| Total AA cost (lower is better) | $2,731.91 | $108.15 | −96.0% GPT-5.6 Luna (High) is 96.0% cheaper |
| End-to-end time (lower is better) | 12.85s | 13.41s | +4.4% GPT-5.6 Luna (High) is 4.4% slower |
| Price per 1M tokens (in / out) | $5.00 / $25.0 | $0.20 / $1.20 | |
| Input modalities | Text, Image | Text, Image |
GPT-5.6 Luna (High) vs Claude Opus 5 (Medium): 28.3% less intelligence, 96.0% less cost to run the AA benchmark, and 4.4% slower end-to-end.
Today's Pareto frontier
Against the frontier
- Claude Opus 5 (Medium) is off the frontier: at the same or lower cost, Claude Opus 5.5 (High, Default Fallback) scores 53.60 for $2172.
- GPT-5.6 Luna (High) is off the frontier: at the same or lower cost, Claude Haiku 5.5 (High) scores 37.80 for $94.1.
Claude Opus 5 (Medium)GPT-5.6 Luna (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/claude-opus-5-medium/gpt-5-6-luna-high · Markdown: /api/v1/compare/claude-opus-5-medium/gpt-5-6-luna-high?format=md · llms.txt · agent skill