We needed to see where our AI budget was going. So we built this.
Internal dev tool we built to track our own AI API costs across Anthropic, OpenAI, and Google. Pulls billing data into a single dashboard with model-level breakdowns, budget alerts, and anomaly detection — the kind of custom data layer any team using multiple APIs could use.
MTD Spend
$0.00
-12%
Daily Burn
$0/d
-8%
Projected
$0.00
on track
Budget
0.0%
$2,500 limit
Provider Distribution
The Problem
Three providers, three dashboards, no idea what you're spending.
Dev teams now routinely use 2-3+ AI providers — Anthropic for complex reasoning, OpenAI for embeddings, Google for multimodal tasks. Each provider has its own billing dashboard, pricing model, and usage API. There's no single source of truth for total API spend.
Cost surprises are the norm. A runaway batch job using an expensive model can burn through hundreds of dollars before anyone notices. Monthly invoices arrive with line items that are hard to trace back to specific projects or teams. Budget planning is guesswork.
Model pricing changes frequently, and teams have no systematic way to evaluate whether they're using the most cost-effective model for each task. The cheapest model that meets quality requirements often isn't the one being used.
The Solution
Unified spend visibility with AI-powered cost intelligence.
We connected our provider API keys once, and the platform pulls usage data automatically. All costs are normalized into a unified view — total spend, per-provider breakdown, per-model comparison, and daily burn rate, all in real time.
The AI layer monitors spending patterns and flags anomalies: unusual spikes, new models appearing, budget threshold breaches, and weekend activity that might indicate unintended cron jobs. Budget gauges show exactly where things stand relative to set limits.
AI-powered reports generate cost optimization recommendations — suggesting cheaper models for specific use cases, identifying low-utilization providers, and projecting month-end spend based on current trajectory.
What it does.
Multi-Provider Spend Tracking
Unified dashboard connecting Anthropic, OpenAI, Google AI, and more via their billing APIs. See total spend, daily burn rate, and budget utilization at a glance.
Model-Level Cost Breakdown
Compare cost per request, cost per 1K tokens, and total spend across every model in use — ranked by efficiency and volume.
Anomaly Detection & Alerts
AI-powered detection of spending spikes, unusual patterns, and budget threshold breaches with real-time alerts.
Budget Management
Set monthly budgets per provider or globally. Visual gauges show utilization, and alerts fire before limits are exceeded.
AI Cost Optimization
Automated recommendations for model switching, usage optimization, and cost reduction based on actual usage patterns.
Natural Language Querying
Ask questions about spend in plain English — 'What's the biggest cost driver?' or 'How does this month compare to last?'