Unit Economics & LLM Inference Cost Sensitivity Model

Constructs a financial sensitivity model calculating token costs, caching efficiencies, and margin thresholds.

Build a comprehensive LLM inference unit economics model based on the following application metrics:

Inputs:
- Monthly Active Users (MAU)
- Daily Queries per User
- Average Prompt Input Tokens & Completion Output Tokens
- Target Latency SLA (p95)

Deliverables:
1. Base monthly token cost across Claude 3.7 Sonnet, GPT-4o, and DeepSeek-V3.
2. Cost reduction impact with 60% prompt caching hit rate.
3. Breakeven subscription price per seat to achieve a 75% gross margin.

Metrics:
[INSERT PARAMETERS HERE]