GPT-6 Astra (Medium) vs Hy3
Which is better for the money: GPT-6 Astra (Medium) by OpenAI or Hy3 by Tencent? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-25.
Trade-off
Verdict
Depends on your budget
GPT-6 Astra (Medium) scores 24.3 points higher (49.6 vs 25.3) but costs 16× as much to run the benchmark ($2434 vs $150; flagship vs mid-range cost classes). They are not alternatives at one budget: pick GPT-6 Astra (Medium) when you need intelligence above 25.3, and Hy3 when 25.3 is enough.
| Metric | GPT-6 Astra (Medium) | Hy3 | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 49.6 | 25.3 | −49.0% Hy3 is 49.0% less intelligent |
| Total AA cost (lower is better) | $2,434.12 | $149.61 | −93.9% Hy3 is 93.9% cheaper |
| End-to-end time (lower is better) | 13.22s | 33.13s | +150.6% Hy3 is 150.6% slower |
| Price per 1M tokens (in / out) | $10.0 / $50.0 | $0.14 / $0.56 | |
| Input modalities | Text, Image | Text |
Hy3 vs GPT-6 Astra (Medium): 49.0% less intelligence, 93.9% less cost to run the AA benchmark, and 150.6% slower end-to-end.
Today's Pareto frontier
Against the frontier
- GPT-6 Astra (Medium) is off the frontier: at the same or lower cost, Claude Opus 5.5 (High, Default Fallback) scores 53.60 for $2172.
- Hy3 is off the frontier: at the same or lower cost, GPT-6 Luna (Max) scores 38.10 for $122.
GPT-6 Astra (Medium)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-6-astra-medium/hy3 · Markdown: /api/v1/compare/gpt-6-astra-medium/hy3?format=md · llms.txt · agent skill