At 10 requests/user/day (totaling 750,000 queries/mo), running RAG / search-augmented answers on Yi-Lightning (01.AI) costs $892.50 / month ($0.357 / user / mo).
RAG / search-augmented answers for a 2,500 active users (series a product) audience means 25,000 requests per day on average, each carrying 8,000 input and 500 output tokens at a 50% cacheable share — so the effective per-request cost on Yi-Lightning (01.AI) is $0.00119, of which $0.00/mo comes back from prompt caching. Per active user that's $0.357/mo — the number to hold against your pricing tier and gross margin. Charging at least $1.785/user/mo keeps gross margin at or above 80%.
Quick answer
For 2,500 Active Users, RAG / search-augmented answers on Yi-Lightning (01.AI) is modeled at 750,000 requests per month, costing $892.50 per month or $0.357 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 750,000 queries.
Direct inference cost per MAU to model your SaaS pricing tiers and gross margins.
Savings generated by exploiting 50% cache hits on prompt context.
| Model | Provider | Monthly Spend (2,500 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| Yi-Lightning (01.AI) (Current) | 01ai | $892.50 | $0.357 | $10,710.00 | 16,384 |
| GPT-5.6 Luna | openai | $1,110.00 | $0.444 | $13,320.00 | 1,050,000 |
| GPT-5.4 mini | openai | $4,162.50 | $1.665 | $49,950.00 | 256,000 |
| GPT-5.4 nano | openai | $1,128.75 | $0.4515 | $13,545.00 | 128,000 |
| GPT-4o mini | openai | $900.00 | $0.36 | $10,800.00 | 128,000 |
| Claude Haiku 4.5 | anthropic | $5,175.00 | $2.07 | $62,100.00 | 1,000,000 |
At 10 queries per user/day with Yi-Lightning (01.AI), the estimated cost is $0.357 per monthly active user (MAU). Total monthly bill for 2,500 Active Users is $892.50.
Assuming a 50% cache hit rate on repeated system and context tokens, prompt caching saves $0.00 per month ($0.00/year) on Yi-Lightning (01.AI).
With a direct COGS of $0.357 per user/month on Yi-Lightning (01.AI), charging at least $1.785 per user/month ensures an 80%+ software gross margin.