Cost Analysis | Geopolitical Trade Pipeline

Cost Per Deep-Dive Run by Provider

Assumes ~2.5M total tokens per deep-dive (1.5M input, 1.0M output). Caching at 70% hit rate for providers that support it.

LLM Pricing Reference

Model	Provider	Input/MTok	Output/MTok	Cached Input	Quality	Use Case
Claude Opus 4.6	Anthropic	$15.0	$75.0	$1.5	Highest	Deep analysis
Claude Sonnet 4.6	Anthropic	$3.0	$15.0	$0.3	High	Standard analysis
GPT-5	OpenAI	$2.5	$10.0	N/A	High	Reasoning tasks
GPT-5.2	OpenAI	$2.5	$14.0	N/A	Highest	Allocation/fund prompts
GPT-4.1 Mini	OpenAI	$0.4	$1.6	$0.1	Low	Triage/filtering
Gemini Flash 3.0	Google	$0.1	$0.4	$0.02	Low	Triage/filtering
Llama 4 Scout	Self-hosted	$0.0	$0.0	$0.0	Medium	Triage (if self-hosted)

Cadence Scenarios

Three operating modes with different cost profiles. The hybrid approach is recommended: cheap triage runs daily, expensive deep-dives only when triggered.

Full Deep-dive Per Event

$30-100/mo

$15-25/event x 2-4 events/month

Every geopolitical event gets the full 12-agent treatment. Highest quality, highest cost.

Triage Only

$15-60/mo

$0.50-2/day continuous scanning

RSS scanning + LLM classification only. Alerts but no deep analysis. Cheapest option.

Hybrid (Recommended)

$50-150/mo

Triage daily + full dive on trigger

Daily triage at ~$1/day. Full deep-dive only when triage scores >= 8. Best cost/quality balance.

Budget ceiling: At $140K AUM, max viable LLM spend is ~$50/day ($1,500/month). All proposed phases are well within this ceiling. Even the most aggressive scenario (Phase 4 autonomous at $150/month) uses only 10% of the budget ceiling.

Monthly Cost Projection

Step-function cost growth across 4 implementation phases. Red dashed line = $50/day budget ceiling ($1,500/month). All phases well below.