DeepSeek V4 Flash 0420 (Max) vs Hy3
Which is better for the money: DeepSeek V4 Flash 0420 (Max) by DeepSeek 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
Hy3 scores 1.1 points higher (25.3 vs 24.2) but costs 1.0× as much to run the benchmark ($150 vs $149; both mid-range class). Pick Hy3 when the extra intelligence is worth the premium, otherwise DeepSeek V4 Flash 0420 (Max).
| Metric | DeepSeek V4 Flash 0420 (Max) | Hy3 | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 24.2 | 25.3 | +4.5% Hy3 is 4.5% more intelligent |
| Total AA cost (lower is better) | $149.48 | $149.61 | +0.1% Hy3 is 0.1% more expensive |
| End-to-end time (lower is better) | 116.79s | 33.29s | −71.5% Hy3 is 71.5% faster |
| Price per 1M tokens (in / out) | $0.11 / $0.23 | $0.14 / $0.56 | |
| Input modalities | Text | Text |
Hy3 vs DeepSeek V4 Flash 0420 (Max): 4.5% more intelligence, 0.1% more cost to run the AA benchmark, and 71.5% faster end-to-end.
Today's Pareto frontier
Against the frontier
- DeepSeek V4 Flash 0420 (Max) is off the frontier: at the same or lower cost, GPT-6 Luna (Max) scores 38.10 for $122.
- Hy3 is off the frontier: at the same or lower cost, GPT-6 Luna (Max) scores 38.10 for $122.
DeepSeek V4 Flash 0420 (Max)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/deepseek-v4-flash-0420/hy3 · Markdown: /api/v1/compare/deepseek-v4-flash-0420/hy3?format=md · llms.txt · agent skill