At 8 requests/user/day (totaling 24,000,000 queries/mo), running Agentic workflow on DeepSeek V4 Flash costs $204,096.00 / month ($2.041 / user / mo).
Agentic workflow for a 100,000 active users (high-volume consumer / enterprise) audience means 800,000 requests per day on average, each carrying 40,000 input and 1,500 output tokens at a 65% cacheable share — so the effective per-request cost on DeepSeek V4 Flash is $0.008504, of which $265,824.00/mo comes back from prompt caching. Per active user that's $2.041/mo — the number to hold against your pricing tier and gross margin. Charging at least $10.205/user/mo keeps gross margin at or above 80%.
Quick answer
For 100,000 Active Users, Agentic workflow on DeepSeek V4 Flash is modeled at 24,000,000 requests per month, costing $204,096.00 per month or $2.041 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 100,000 Active Users generating 24,000,000 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 (100,000 Active Users) | Cost / User / Mo | Annual Spend | Context Limit |
|---|---|---|---|---|---|
| DeepSeek V4 Flash (Current) | deepseek | $204,096.00 | $2.041 | $2,449,152.00 | 1,000,000 |
| GPT-5.6 Luna | openai | $122,880.00 | $1.229 | $1,474,560.00 | 1,050,000 |
| GPT-5.4 mini | openai | $460,800.00 | $4.608 | $5,529,600.00 | 256,000 |
| GPT-5.4 nano | openai | $124,680.00 | $1.247 | $1,496,160.00 | 128,000 |
| GPT-4o mini | openai | $118,800.00 | $1.188 | $1,425,600.00 | 128,000 |
| Claude Haiku 4.5 | anthropic | $578,400.00 | $5.784 | $6,940,800.00 | 1,000,000 |
At 8 queries per user/day with DeepSeek V4 Flash, the estimated cost is $2.041 per monthly active user (MAU). Total monthly bill for 100,000 Active Users is $204,096.00.
Assuming a 65% cache hit rate on repeated system and context tokens, prompt caching saves $265,824.00 per month ($3,189,888.00/year) on DeepSeek V4 Flash.
With a direct COGS of $2.041 per user/month on DeepSeek V4 Flash, charging at least $10.205 per user/month ensures an 80%+ software gross margin.