Simulating realistic Document summarization parameters (25,000 in / 600 out with 10% cache reuse). Codestral 2508 (OpenRouter) delivers a 88% cost reduction over Cohere Command A+.
| Traffic Volume Tier | Cohere Command A+ Monthly | Codestral 2508 (OpenRouter) Monthly | Monthly Savings by picking Codestral 2508 (OpenRouter) |
|---|---|---|---|
| 1,000 reqs/mo (Dev/Testing) | $62.875 | $7.365 | Save $55.51 / mo |
| 10,000 reqs/mo (Small App) | $628.75 | $73.65 | Save $555.10 / mo |
| 100,000 reqs/mo (Growth Production) | $6,287.50 | $736.50 | Save $5,551.00 / mo |
| 1,000,000 reqs/mo (Scale SaaS) | $62,875.00 | $7,365.00 | Save $55,510.00 / mo |
Codestral 2508 (OpenRouter) is 88% cheaper for Document summarization workloads. At standard Document summarization parameter ratios (25,000 input tokens, 600 output tokens, 10% cache hit), Codestral 2508 (OpenRouter) costs $0.007365 per request compared to $0.0629 on Cohere Command A+.
Cohere Command A+ offers a context window of 128,000 tokens (max output: 8,192), while Codestral 2508 (OpenRouter) offers 256,000 tokens (max output: 204,800).
At 100,000 requests per month, using Codestral 2508 (OpenRouter) saves $5,551.00 every month (or $66,612.00 annually) compared to Cohere Command A+.
Long-context models pay off here — compare price per 1M tokens at your true document size. Summarize once, store the result; don't re-summarize unchanged documents. For batch backfills, nightly jobs can use cache-friendly request ordering.