GPT-5.6 Luna (Xhigh) vs Hy3
Which is better for the money: GPT-5.6 Luna (Xhigh) by OpenAI or Hy3 by Tencent? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-24.
Trade-off
Verdict
Depends on your budget
GPT-5.6 Luna (Xhigh) scores 9.3 points higher (34.6 vs 25.3) but costs 1.2× as much to run the benchmark ($179 vs $150; both mid-range class). Pick GPT-5.6 Luna (Xhigh) when the extra intelligence is worth the premium, otherwise Hy3.
| Metric | GPT-5.6 Luna (Xhigh) | Hy3 | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 34.6 | 25.3 | −26.9% Hy3 is 26.9% less intelligent |
| Total AA cost (lower is better) | $179.33 | $149.61 | −16.6% Hy3 is 16.6% cheaper |
| End-to-end time (lower is better) | 30.83s | 33.29s | +8.0% Hy3 is 8.0% slower |
| Price per 1M tokens (in / out) | $0.20 / $1.20 | $0.14 / $0.56 | |
| Input modalities | Text, Image | Text |
Hy3 vs GPT-5.6 Luna (Xhigh): 26.9% less intelligence, 16.6% less cost to run the AA benchmark, and 8.0% slower end-to-end.
Today's Pareto frontier
Against the frontier
- GPT-5.6 Luna (Xhigh) is off the frontier: at the same or lower cost, Claude Haiku 5.5 (Xhigh) scores 41.20 for $158.
- Hy3 is off the frontier: at the same or lower cost, GPT-6 Luna (Max) scores 38.10 for $122.
GPT-5.6 Luna (Xhigh)Hy3Today'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-xhigh/hy3 · Markdown: /api/v1/compare/gpt-5-6-luna-xhigh/hy3?format=md · llms.txt · agent skill