At 40 requests/user/day (totaling 120,000 queries/mo), running AI coding assistant on Llama 3.2 3B Instruct costs $55.20 / month ($0.552 / user / mo).
Total expected API invoice for 100 Active Users generating 120,000 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 60% cache hits on prompt context.
| Model | Provider | Monthly Spend (100 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Llama 3.2 3B Instruct (Current) | meta | $55.20 | $0.552 | $662.40 | 128,000 |
| GPT-5.6 Luna | openai | $420.48 | $4.205 | $5,045.76 | 1,050,000 |
| GPT-5.4 mini | openai | $1,576.80 | $15.768 | $18,921.60 | 256,000 |
| GPT-5.4 nano | openai | $432.48 | $4.325 | $5,189.76 | 128,000 |
| GPT-4o mini | openai | $295.20 | $2.952 | $3,542.40 | 128,000 |
| Claude Haiku 4.5 | anthropic | $1,862.40 | $18.624 | $22,348.80 | 1,000,000 |
At 40 queries per user/day with Llama 3.2 3B Instruct, the estimated cost is $0.552 per monthly active user (MAU). Total monthly bill for 100 Active Users is $55.20.
Assuming a 60% cache hit rate on repeated system and context tokens, prompt caching saves $0.00 per month ($0.00/year) on Llama 3.2 3B Instruct.
With a direct COGS of $0.552 per user/month on Llama 3.2 3B Instruct, charging at least $2.76 per user/month ensures an 80%+ software gross margin.