Cost Calculator
Estimate AI processing, data, and infrastructure costs at any scale — with live platform data
Loading Live Platform Data...
Projected Scale
Scale at County
Meeting summaries, conflict checks, proposal analysis, chat, research synthesis, sentiment analysis
ArcGIS geocoding, NetFile sync, CivicPlus agendas (most are free)
1 server(s) running tRPC API, subdomain routing, background jobs
Content delivery, static assets, API responses to end users
Storage + request units for all jurisdiction data, officers, contributions, proceedings
Documents, images, meeting recordings, Form 700 PDFs, proposal attachments
Full-text search across officers, legislation, meetings, proposals
Revenue vs. Cost Analysis
| Scale | Jurisdictions | Monthly Cost | Avg Contract | Monthly Revenue | Gross Margin |
|---|---|---|---|---|---|
| Single City | 1 | $68 | $2.3K/mo | $2.3K | 97% |
| County | 19 | $677 | $2.1K/mo | $20.8K | 97% |
| State (CA) | 540 | $19.9K | $2.2K/mo | $178.0K | 89% |
| Region (5 States) | 2.8K | $98.0K | $2.1K/mo | $590.2K | 83% |
| National (50 States) | 22.6K | $799.5K | $2.1K/mo | $2.4M | 67% |
AI Processing Cost Detail
| AI Function | Cost/Call | Est. Calls/mo | Monthly Cost |
|---|---|---|---|
| Meeting Summaries | $0.12 | 136 | $16.32 |
| Conflict Checks | $0.04 | 1.0K | $40.40 |
| Proposal Analysis | $0.35 | 260 | $91.00 |
| Chat Messages | $0.02 | 15.8K | $316.00 |
| Research Synthesis | $0.50 | 165 | $82.50 |
| Sentiment Analysis | $0.03 | 690 | $20.70 |
| Budget Analysis | $0.08 | 86 | $6.88 |
Scale Comparison — All Tiers
Assumptions & Methodology
City Size Tiers
- Small <50K population — 4 meetings/mo, 5 officers, 50 MAU
- Medium 50K–200K — 8 meetings/mo, 15 officers, 200 MAU
- Large 200K+ — 16 meetings/mo, 40 officers, 800 MAU
LLM Pricing
Based on GPT-4o-class models at current API pricing. Costs decrease ~30% annually as models improve and competition increases. Batch processing and caching can reduce costs by 40–60%.
Live Data
The Live Platform Data panel queries all 96 database tables in real-time to show actual entity counts, AI invocation estimates, and storage usage. AI costs are derived from entity counts multiplied by per-call pricing.
Infrastructure
TiDB Serverless (auto-scaling MySQL), S3 for file storage, CDN for static assets. Compute based on t3.medium equivalent at ~$34/mo. One server per ~500 jurisdictions.
Cost Optimization Strategies
- • LLM response caching for repeated queries (est. 40% savings)
- • Batch processing for non-real-time analysis (est. 30% savings)
- • Smaller models for simple tasks (sentiment, classification)
- • Reserved capacity pricing at scale (est. 20% savings)
- • Edge caching for static compliance data