DeepSeek V3 vs Qwen2.5 Coder 32B Instruct
Side by side from live data: prices, limits, who serves each model and how reliably, and how they do on our chess arena.
In short
- DeepSeek V3 is 1.7× cheaper at a typical mix of 3 input tokens per output token.
- DeepSeek V3 accepts longer prompts: 164K vs 33K tokens.
- DeepSeek V3 is the newer model (released 2024-12-26).
| DeepSeek V3 | Qwen2.5 Coder 32B Instruct | |
|---|---|---|
| Vendor | DeepSeek | qwen |
| Released | 2024-12-26 | 2024-11-11 |
| Input / 1M | $0.257 | $0.66 |
| Output / 1M | $1.03 | $1.00 |
| Cached input / 1M | — | — |
| Blended / 1M (3:1) | $0.45 | $0.745 |
| Context | 164K | 33K |
| Max output | 16K | 29K |
| Accepts | text | text |
| Reasoning | No | No |
| Providers on OpenRouter | 3 | 1 |
| Cheapest provider, output | $0.89 | $1.00 |
| Best provider uptime, 24 h | 99.51% | 100% |
| Chess arena rating | not playing | not playing |
| Vendor API status | Operational | Operational |
Cost calculator
DeepSeek V3
$30.87 per month$1.03 per day
Qwen2.5 Coder 32B Instruct
$54.60 per month$1.82 per day
DeepSeek V3 saves $23.73 a month on this workload.
At list prices per 1M tokens; prompt caching and batch discounts are not included.
Compare DeepSeek V3 with
Compare Qwen2.5 Coder 32B Instruct with