At 200 requests/user/day (totaling 6,000,000 queries/mo), running Data extraction & tagging on Codestral 2501 costs $4,302.00 / month ($4.302 / user / mo).
Total expected API invoice for 1,000 Active Users generating 6,000,000 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 20% cache hits on prompt context.
| Model | Provider | Monthly Spend (1,000 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Codestral 2501 (Current) | mistral | $4,302.00 | $4.302 | $51,624.00 | 256,000 |
| GPT-5.3 Codex | openai | $38,220.00 | $38.22 | $458,640.00 | 256,000 |
| DeepSeek Coder V2.5 | deepseek | $1,797.60 | $1.798 | $21,571.20 | 128,000 |
| Grok Build 0.1 | xai | $13,080.00 | $13.08 | $156,960.00 | 256,000 |
| Qwen 2.5 Coder 32B | qwen | $2,868.00 | $2.868 | $34,416.00 | 128,000 |
| DeepInfra — Qwen 2.5 Coder 32B | deepinfra | $1,320.00 | $1.32 | $15,840.00 | 128,000 |
At 200 queries per user/day with Codestral 2501, the estimated cost is $4.302 per monthly active user (MAU). Total monthly bill for 1,000 Active Users is $4,302.00.
Assuming a 20% cache hit rate on repeated system and context tokens, prompt caching saves $648.00 per month ($7,776.00/year) on Codestral 2501.
With a direct COGS of $4.302 per user/month on Codestral 2501, charging at least $21.51 per user/month ensures an 80%+ software gross margin.