At 6 requests/user/day (totaling 18,000 queries/mo), running Customer support chatbot on Groq LPU — Llama 3.1 8B Instant costs $3.654 / month ($0.0365 / user / mo).
Total expected API invoice for 100 Active Users generating 18,000 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 70% cache hits on prompt context.
| Model | Provider | Monthly Spend (100 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Groq LPU — Llama 3.1 8B Instant (Current) | groq | $3.654 | $0.0365 | $43.848 | 128,000 |
| GPT-5.6 Luna | openai | $12.222 | $0.1222 | $146.664 | 1,050,000 |
| GPT-5.4 mini | openai | $45.832 | $0.4583 | $549.99 | 256,000 |
| GPT-5.4 nano | openai | $12.537 | $0.1254 | $150.444 | 128,000 |
| GPT-4o mini | openai | $9.923 | $0.0992 | $119.07 | 128,000 |
| Claude Haiku 4.5 | anthropic | $54.81 | $0.5481 | $657.72 | 1,000,000 |
At 6 queries per user/day with Groq LPU — Llama 3.1 8B Instant, the estimated cost is $0.0365 per monthly active user (MAU). Total monthly bill for 100 Active Users is $3.654.
Assuming a 70% cache hit rate on repeated system and context tokens, prompt caching saves $0.00 per month ($0.00/year) on Groq LPU — Llama 3.1 8B Instant.
With a direct COGS of $0.0365 per user/month on Groq LPU — Llama 3.1 8B Instant, charging at least $0.1827 per user/month ensures an 80%+ software gross margin.