GPT-5.6 Luna (Non-reasoning) vs gpt-oss-20b (High)
Which is better for the money: GPT-5.6 Luna (Non-reasoning) by OpenAI or gpt-oss-20b (High) by OpenAI? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-26.
Clear winner
Verdict
Pick GPT-5.6 Luna (Non-reasoning)
gpt-oss-20b (High) vs GPT-5.6 Luna (Non-reasoning): 41.9% less intelligence, 25.4% more cost to run the AA benchmark, and 189.6% slower end-to-end.
| Metric | GPT-5.6 Luna (Non-reasoning) | gpt-oss-20b (High) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 15.5 | 9.0 | −41.9% gpt-oss-20b (High) is 41.9% less intelligent |
| Total AA cost (lower is better) | $22.97 | $28.81 | +25.4% gpt-oss-20b (High) is 25.4% more expensive |
| End-to-end time (lower is better) | 4.98s | 14.42s | +189.6% gpt-oss-20b (High) is 189.6% slower |
| Price per 1M tokens (in / out) | $0.20 / $1.20 | $0.05 / $0.22 | |
| Input modalities | Text, Image | Text |
Today's Pareto frontier
Against the frontier
- GPT-5.6 Luna (Non-reasoning) is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
- gpt-oss-20b (High) is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
GPT-5.6 Luna (Non-reasoning)gpt-oss-20b (High)Today's Pareto frontierOpen-weights frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/gpt-5-6-luna-non-reasoning/gpt-oss-20b · Markdown: /api/v1/compare/gpt-5-6-luna-non-reasoning/gpt-oss-20b?format=md · llms.txt · agent skill