README.md

---

title: Dispatch AI — AI Cost Calculator
emoji: 💰
colorFrom: blue
colorTo: indigo
sdk: gradio
sdk_version: "4.44.1"
app_file: app.py
pinned: true
license: other
tags:
  - cost-calculator
  - on-device
  - savings
  - co2
  - dispatch-ai
---


# Dispatch AI — AI Cost Calculator 💰

Input your cloud API usage (tokens/month) → see cost savings by switching to on-device AI inference. Shows breakeven point, 1-year savings, and CO2 reduction.

Built by **Dispatch AI (FZE)** — UAE · License 10818.

## Features

- 💵 **Total savings** — cloud vs on-device over your projection period
- ⏱️ **Breakeven point** — when on-device becomes cheaper than cloud
- 🌍 **CO2 reduction** — environmental impact of switching
- 📊 **Side-by-side comparison** — monthly cost, total cost, CO2 emissions
- 🌳 **Tree equivalent** — how many trees your CO2 savings represent
- ⚡ **Capacity check** — can your phone fleet handle the token volume?
- 📋 **8 cloud providers** — GPT-4o, GPT-4o-mini, Claude 3.5 Sonnet/Haiku, Llama, Mistral, Gemini
- 🌙 **Dark theme** (#0A0F1A bg, #1FE0E6 cyan accent)

## How It Works

1. Enter your monthly input/output token usage (in millions)
2. Pick your current cloud provider
3. Set how many phones you'd deploy
4. Choose projection period (1-36 months)
5. Click **Calculate Savings**

## Assumptions

- Phone: Samsung S20 FE ($400, 24-month lifespan)
- Inference: 16 t/s (Qwen2.5-1.5B Q4_K_M on SD865)
- Electricity: ~$2.50/month/phone
- CO2: 0.4 kg/1M tokens (cloud) vs 0.384 kg/month (phone)

## License

© 2026 Dispatch AI FZE — Sharjah, UAE · License 10818. All rights reserved.