At 10 requests/user/day (totaling 300,000 queries/mo), running RAG / search-augmented answers on DeepSeek V4 Flash costs $742.80 / month ($0.7428 / user / mo).
Total expected API invoice for 1,000 Active Users generating 300,000 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 50% cache hits on prompt context.
| Model | Provider | Monthly Spend (1,000 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| DeepSeek V4 Flash (Current) | deepseek | $742.80 | $0.7428 | $8,913.60 | 1,000,000 |
| GPT-5.6 Luna | openai | $444.00 | $0.444 | $5,328.00 | 1,050,000 |
| GPT-5.4 mini | openai | $1,665.00 | $1.665 | $19,980.00 | 256,000 |
| GPT-5.4 nano | openai | $451.50 | $0.4515 | $5,418.00 | 128,000 |
| GPT-4o mini | openai | $360.00 | $0.36 | $4,320.00 | 128,000 |
| Claude Haiku 4.5 | anthropic | $2,070.00 | $2.07 | $24,840.00 | 1,000,000 |
At 10 queries per user/day with DeepSeek V4 Flash, the estimated cost is $0.7428 per monthly active user (MAU). Total monthly bill for 1,000 Active Users is $742.80.
Assuming a 50% cache hit rate on repeated system and context tokens, prompt caching saves $511.20 per month ($6,134.40/year) on DeepSeek V4 Flash.
With a direct COGS of $0.7428 per user/month on DeepSeek V4 Flash, charging at least $3.714 per user/month ensures an 80%+ software gross margin.