At 15 requests/user/day (totaling 4,500 queries/mo), running Translation on Text Embedding 3 (Large) costs $2.925 / month ($0.2925 / user / mo).
Total expected API invoice for 10 Active Users generating 4,500 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 15% cache hits on prompt context.
| Model | Provider | Monthly Spend (10 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Text Embedding 3 (Large) (Current) | openai | $2.925 | $0.2925 | $35.10 | 8,191 |
| Text Embedding 3 (Small) | openai | $0.45 | $0.045 | $5.40 | 8,191 |
| Text Embedding 004 | $0.45 | $0.045 | $5.40 | 8,192 |
At 15 queries per user/day with Text Embedding 3 (Large), the estimated cost is $0.2925 per monthly active user (MAU). Total monthly bill for 10 Active Users is $2.925.
Assuming a 15% cache hit rate on repeated system and context tokens, prompt caching saves $0.00 per month ($0.00/year) on Text Embedding 3 (Large).
With a direct COGS of $0.2925 per user/month on Text Embedding 3 (Large), charging at least $1.463 per user/month ensures an 80%+ software gross margin.