GPT-5.6 Luna (Low) vs gpt-oss-20b (High)
Which is better for the money: GPT-5.6 Luna (Low) 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 (Low)
gpt-oss-20b (High) vs GPT-5.6 Luna (Low): 57.1% less intelligence, 6.6% more cost to run the AA benchmark, and 144.4% slower end-to-end.
| Metric | GPT-5.6 Luna (Low) | gpt-oss-20b (High) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 21.0 | 9.0 | −57.1% gpt-oss-20b (High) is 57.1% less intelligent |
| Total AA cost (lower is better) | $27.02 | $28.81 | +6.6% gpt-oss-20b (High) is 6.6% more expensive |
| End-to-end time (lower is better) | 5.90s | 14.42s | +144.4% gpt-oss-20b (High) is 144.4% 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 (Low) 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 (Low)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-low/gpt-oss-20b · Markdown: /api/v1/compare/gpt-5-6-luna-low/gpt-oss-20b?format=md · llms.txt · agent skill