At 6 requests/user/day (totaling 450,000 queries/mo), running Customer support chatbot on Claude 3.5 Haiku costs $1,096.20 / month ($0.4385 / user / mo).
Customer support chatbot for a 2,500 active users (series a product) audience means 15,000 requests per day on average, each carrying 3,500 input and 350 output tokens at a 70% cacheable share — so the effective per-request cost on Claude 3.5 Haiku is $0.002436, of which $793.80/mo comes back from prompt caching. Per active user that's $0.4385/mo — the number to hold against your pricing tier and gross margin. Charging at least $2.192/user/mo keeps gross margin at or above 80%.
Quick answer
For 2,500 Active Users, Customer support chatbot on Claude 3.5 Haiku is modeled at 450,000 requests per month, costing $1,096.20 per month or $0.4385 per active user.
Method & trust
This projection combines the selected active-user tier, requests per user per day, input/output tokens per request, and the modeled cache share. It is an operating estimate, not a provider invoice; production traffic and retries should be measured against it.
Total expected API invoice for 2,500 Active Users generating 450,000 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 (2,500 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Claude 3.5 Haiku (Current) | anthropic | $1,096.20 | $0.4385 | $13,154.40 | 200,000 |
| GPT-5.6 Luna | openai | $305.55 | $0.1222 | $3,666.60 | 1,050,000 |
| GPT-5.4 mini | openai | $1,145.813 | $0.4583 | $13,749.75 | 256,000 |
| GPT-5.4 nano | openai | $313.425 | $0.1254 | $3,761.10 | 128,000 |
| GPT-4o mini | openai | $248.063 | $0.0992 | $2,976.75 | 128,000 |
| Claude Haiku 4.5 | anthropic | $1,370.25 | $0.5481 | $16,443.00 | 1,000,000 |
At 6 queries per user/day with Claude 3.5 Haiku, the estimated cost is $0.4385 per monthly active user (MAU). Total monthly bill for 2,500 Active Users is $1,096.20.
Assuming a 70% cache hit rate on repeated system and context tokens, prompt caching saves $793.80 per month ($9,525.60/year) on Claude 3.5 Haiku.
With a direct COGS of $0.4385 per user/month on Claude 3.5 Haiku, charging at least $2.192 per user/month ensures an 80%+ software gross margin.