DeepSeek V4 Pro 0423 vs Qwen3 235B A22B Thinking 2507
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
- Qwen3 235B A22B Thinking 2507 is 1.6× cheaper at a typical mix of 3 input tokens per output token.
- DeepSeek V4 Pro 0423 accepts longer prompts: 1M vs 128K tokens.
- DeepSeek V4 Pro 0423 output costs $1.50 per 1M via Relace, against the list price of $3.20.
- DeepSeek V4 Pro 0423 is the newer model (released 2026-04-24).
| DeepSeek V4 Pro 0423 | Qwen3 235B A22B Thinking 2507 | |
|---|---|---|
| Vendor | DeepSeek | Qwen |
| Released | 2026-04-24 | 2025-07-25 |
| Input / 1M | $1.60 | $0.45 |
| Output / 1M | $3.20 | $3.50 |
| Cached input / 1M | $0.135 | — |
| Blended / 1M (3:1) | $2.00 | $1.21 |
| Context | 1M | 128K |
| Max output | 393K | 16K |
| Accepts | text | text |
| Reasoning | Yes | Yes |
| Providers on OpenRouter | 15 | 1 |
| Cheapest provider, output | $1.50 | $3.50 |
| Best provider uptime, 24 h | 100% | 100% |
| Chess arena rating | not playing | not playing |
| Vendor API status | Operational | Operational |
Cost calculator
DeepSeek V4 Pro 0423
$144 per month$4.80 per day
Qwen3 235B A22B Thinking 2507
$79.50 per month$2.65 per day
Qwen3 235B A22B Thinking 2507 saves $64.50 a month on this workload.
At list prices per 1M tokens; prompt caching and batch discounts are not included.
Compare DeepSeek V4 Pro 0423 with
Compare Qwen3 235B A22B Thinking 2507 with