Mistral Devstral 2 2512 vs Qwen3 VL 235B A22B 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
- Qwen3 VL 235B A22B Instruct is 1.3× cheaper at a typical mix of 3 input tokens per output token.
- Both accept prompts up to 262K tokens.
- Qwen3 VL 235B A22B Instruct output costs $0.88 per 1M via DeepInfra, against the list price of $1.90.
- Devstral 2 2512 is the newer model (released 2025-12-09).
| Devstral 2 2512 | Qwen3 VL 235B A22B Instruct | |
|---|---|---|
| Vendor | Mistral | Qwen |
| Released | 2025-12-09 | 2025-09-23 |
| Input / 1M | $0.40 | $0.21 |
| Output / 1M | $2.00 | $1.90 |
| Cached input / 1M | $0.04 | $0.10 |
| Blended / 1M (3:1) | $0.80 | $0.632 |
| Context | 262K | 262K |
| Max output | 210K | 33K |
| Accepts | text, file | text, image |
| Reasoning | No | No |
| Providers on OpenRouter | 1 | 3 |
| Cheapest provider, output | $2.00 | $0.88 |
| Best provider uptime, 24 h | 99.68% | 100% |
| Chess arena rating | not playing | not playing |
| Vendor API status | Operational | Operational |
Cost calculator
Devstral 2 2512
$54.00 per month$1.80 per day
Qwen3 VL 235B A22B Instruct
$41.10 per month$1.37 per day
Qwen3 VL 235B A22B Instruct saves $12.90 a month on this workload.
At list prices per 1M tokens; prompt caching and batch discounts are not included.
Compare Devstral 2 2512 with
Compare Qwen3 VL 235B A22B Instruct with