Side-by-side comparison of pricing, 12 benchmarks, and generation speed.
| Metric | Qwen3 235B A22B (Non-reasoning) | Hermes 4 - Llama-3.1 405B (Reasoning) |
|---|---|---|
| Input ($/M tokens) | $0.7 | $1 |
| Output ($/M tokens) | $2.8 | $3 |
Data from Artificial Analysis API — 12 benchmarks
Qwen3 235B A22B (Non-reasoning) is cheaper overall. Its blended price (3:1 input/output ratio) is $1.23/M tokens vs $1.50/M for Hermes 4 - Llama-3.1 405B (Reasoning).
Hermes 4 - Llama-3.1 405B (Reasoning) wins 7 out of 12 benchmarks compared to 4 for Qwen3 235B A22B (Non-reasoning). See the detailed benchmark chart above for per-category results.
Qwen3 235B A22B (Non-reasoning) generates tokens faster at 64 tok/s vs 38 tok/s. However, Hermes 4 - Llama-3.1 405B (Reasoning) has lower time-to-first-token (0.80s vs 1.26s).
Choose based on your priorities: Qwen3 235B A22B (Non-reasoning) for lower cost, Hermes 4 - Llama-3.1 405B (Reasoning) for stronger benchmark performance, and Qwen3 235B A22B (Non-reasoning) for faster generation. For latency-sensitive apps, check the TTFT comparison above.