Mistral Codestral 2508 vs Qwen2.5 VL 72B 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
- Codestral 2508 is 1.9× cheaper at a typical mix of 3 input tokens per output token.
- Codestral 2508 accepts longer prompts: 256K vs 128K tokens.
- Codestral 2508 is the newer model (released 2025-08-01).
| Codestral 2508 | Qwen2.5 VL 72B Instruct | |
|---|---|---|
| Vendor | Mistral | Qwen |
| Released | 2025-08-01 | 2025-02-01 |
| Input / 1M | $0.30 | $0.80 |
| Output / 1M | $0.90 | $1.00 |
| Cached input / 1M | $0.03 | $0.40 |
| Blended / 1M (3:1) | $0.45 | $0.85 |
| Context | 256K | 128K |
| Max output | 205K | 115K |
| Accepts | text, file | text, image |
| Reasoning | No | No |
| Providers on OpenRouter | 2 | 1 |
| Cheapest provider, output | $0.90 | $1.00 |
| Best provider uptime, 24 h | 100% | 99.93% |
| Chess arena rating | not playing | not playing |
| Vendor API status | Operational | Operational |
Cost calculator
Codestral 2508
$31.50 per month$1.05 per day
Qwen2.5 VL 72B Instruct
$63.00 per month$2.10 per day
Codestral 2508 saves $31.50 a month on this workload.
At list prices per 1M tokens; prompt caching and batch discounts are not included.
Compare Codestral 2508 with
Compare Qwen2.5 VL 72B Instruct with