Files
timmy-config/docs/fleet-cost-report.md
Alexander Whitestone 634a72f288 feat: Force Multipliers #541 velocity tracking + #542 fleet cost report
Adds velocity_tracking() to track burn velocity across repos.
Adds fleet cost report in docs/fleet-cost-report.md.

Fixes #541
Fixes #542
2026-04-06 23:09:48 -04:00

1.8 KiB

Fleet Cost & Resource Inventory

Last audited: 2026-04-06 Owner: Timmy Foundation Ops

Model Inference Providers

Provider Type Cost Model Agents Using Est. Monthly
OpenRouter (qwen3.6-plus:free) API Free tier Code Claw, Timmy $0
OpenRouter (various) API Credits Fleet varies
Anthropic (Claude Code) API Subscription claw-code fallback ~$20/mo
Google AI Studio (Gemini) Portal Free daily quota Strategic tasks $0
Ollama (local) Local Electricity only Mac Hermes $0

VPS Infrastructure

Server IP Cost/Mo Running Key Services
Ezra 143.198.27.163 $12/mo Yes Gitea, agent hosting
Allegro 167.99.126.228 $12/mo Yes Agent hosting
Bezalel 159.203.146.185 $12/mo Yes Evennia, agent hosting
Total VPS ~$36/mo

Local Infrastructure

Resource Cost
MacBook (owner-provided) Electricity only
Ollama models (downloaded) Free
Git/Dev tools (OSS) Free

Cost Recommendations

Agent Verdict Reason
Code Claw (OpenRouter) DEPLOY Free tier, adequate for small patches
Gemini AI Studio DEPLOY Free daily quota, good for heavy reasoning
Ollama local DEPLOY No API cost, sovereignty
VPS fleet DEPLOY $36/mo for 3 servers is minimal
Anthropic subscriptions MONITOR Burn $20/mo per seat; watch usage vs output

Monthly Burn Rate Estimate

  • Floor (essential): ~$36/mo (VPS only)
  • Current (with Anthropic): ~$56-76/mo
  • Ceiling (all providers maxed): ~$100+/mo

Notes

  • No GPU instances provisioned yet (no cloud costs)
  • OpenRouter free tier has rate limits
  • Gemini AI Studio daily quota resets automatically