Qwen2.5 72B Instruct vs DeepSeek-R1
DeepSeek-R1 is cheaper on output tokens, while DeepSeek-R1 offers a larger context window. Choose Qwen2.5 72B Instruct or DeepSeek-R1 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Qwen2.5 72B Instruct | DeepSeek-R1 |
|---|---|---|
| Provider | Alibaba | Azure |
| Input / 1M tokens | $1.40 | $1.35 |
| Output / 1M tokens | $5.60 | $5.40 |
| Context window | 131K | 164K |
| Parameters | — | 671B |
| Open weights | Yes | Yes |
| Released | Sep 2024 | Jan 2025 |
| LMArena (Chatbot Arena) Elo score | 1269 | 1428 |
| GPQA Diamond score | 49.1 | 76.3 |
| Epoch Capabilities Index score | — | 141 |
FAQ
Is Qwen2.5 72B Instruct or DeepSeek-R1 cheaper?
DeepSeek-R1 is cheaper on output tokens ($5.40 vs $5.60 per 1M).
Which has the larger context window, Qwen2.5 72B Instruct or DeepSeek-R1?
DeepSeek-R1 has the larger context window (164K tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.