Side-by-side comparison of pricing, 12 benchmarks, and generation speed.
| Metric | Qwen2.5 Instruct 72B | Hy3-preview (Non-reasoning) |
|---|---|---|
| Input ($/M tokens) | $0.475 | $0.123 |
| Output ($/M tokens) | $0.495 | $0.43 |
Data from Artificial Analysis API — 12 benchmarks
Hy3-preview (Non-reasoning) is cheaper overall. Its blended price (3:1 input/output ratio) is $0.20/M tokens vs $0.48/M for Qwen2.5 Instruct 72B.
Hy3-preview (Non-reasoning) wins 6 out of 12 benchmarks compared to 5 for Qwen2.5 Instruct 72B. See the detailed benchmark chart above for per-category results.
Hy3-preview (Non-reasoning) generates tokens faster at 191 tok/s vs 0 tok/s. Qwen2.5 Instruct 72B also has lower time-to-first-token (0.00s vs 1.85s).
Choose based on your priorities: Hy3-preview (Non-reasoning) for lower cost, Hy3-preview (Non-reasoning) for stronger benchmark performance, and Hy3-preview (Non-reasoning) for faster generation. For latency-sensitive apps, check the TTFT comparison above.