DeepSeek's DeepSeek and Alibaba's Qwen, compared through their current models on the Respan Index. The strongest DeepSeek model today is DeepSeek V4.1 Flash (#20 on the Respan Index, 50.9); the strongest Qwen model is Qwen3.8-Max (#22, 50.3). DeepSeek's best models lead in coding and reasoning; Qwen's lead in agents, knowledge and human preference.
| Category | DeepSeek | Qwen |
|---|---|---|
| coding | DeepSeek V4.1 Flash 40.1 | Qwen3.8-Max 38.9 |
| agents | DeepSeek V4.1 Flash 61.2 | Qwen3.8-Max 68.0 |
| reasoning | DeepSeek V4.1 Flash 48.2 | Qwen3.8-Max 40.8 |
| knowledge | DeepSeek V4 Pro 52.6 | Qwen3.7-Max 55.2 |
| human preference | DeepSeek V4.1 Flash 55.7 | Qwen3.8-Max |
Each family's highest Respan Index category score; the higher one is bold. * Estimated: no published scores in that category. How the Respan Index works
Respan Index rank and score, blended list price per 1M tokens.
Respan Index rank and score, blended list price per 1M tokens.
The cheapest ranked DeepSeek model is DeepSeek V4.1 Flash at $0.525 per 1M tokens blended; the cheapest Qwen model is Qwen3.5-Flash at $0.175.