At 8 requests/user/day (totaling 240,000 queries/mo), running Agentic workflow on Llama 3.2 11B Vision costs $1,593.60 / month ($1.594 / user / mo).
Total expected API invoice for 1,000 Active Users generating 240,000 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 65% cache hits on prompt context.
| Model | Provider | Monthly Spend (1,000 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Llama 3.2 11B Vision (Current) | meta | $1,593.60 | $1.594 | $19,123.20 | 128,000 |
| GPT-5.6 Luna | openai | $1,228.80 | $1.229 | $14,745.60 | 1,050,000 |
| GPT-5.4 mini | openai | $4,608.00 | $4.608 | $55,296.00 | 256,000 |
| GPT-5.4 nano | openai | $1,246.80 | $1.247 | $14,961.60 | 128,000 |
| GPT-4o mini | openai | $1,188.00 | $1.188 | $14,256.00 | 128,000 |
| Claude Haiku 4.5 | anthropic | $5,784.00 | $5.784 | $69,408.00 | 1,000,000 |
At 8 queries per user/day with Llama 3.2 11B Vision, the estimated cost is $1.594 per monthly active user (MAU). Total monthly bill for 1,000 Active Users is $1,593.60.
Assuming a 65% cache hit rate on repeated system and context tokens, prompt caching saves $0.00 per month ($0.00/year) on Llama 3.2 11B Vision.
With a direct COGS of $1.594 per user/month on Llama 3.2 11B Vision, charging at least $7.968 per user/month ensures an 80%+ software gross margin.