Overview
Token Tracking monitors your AI token usage and costs across all providers in real time. See exactly how much you’re spending, set budget alerts, and export usage data.
Supported Providers
| Provider | Models |
|---|---|
| Anthropic | Claude Sonnet 4, Claude Opus 4, Haiku |
| OpenAI | GPT-4o, GPT-4, GPT-3.5 |
| Gemini Pro, Gemini Ultra | |
| Mistral | Mistral Large, Medium, Small |
| Groq | LLaMA, Mixtral |
| Local | Ollama, LM Studio (free) |
Dashboard Tabs
Overview
The main dashboard shows:
- Summary cards — Total tokens, total cost, request count, active sessions
- Cost trend chart — Line chart showing spending over time
- Provider breakdown — Pie chart of usage by provider
- Model usage table — Per-model token counts, costs, and request counts
- Top requests — Table of highest-cost individual API calls
Pricing
Configure per-model pricing rates:
- Default rates are pre-loaded for all supported providers
- Edit input/output cost per 1K tokens for any model
- Reset to defaults at any time
Budgets
Set up spending alerts to avoid surprises:
- Create alerts with daily, weekly, monthly, or total budget limits
- Warning thresholds — Get notified at 80% and 100% of your budget
- Per-project budgets — Different limits for different projects
Real-Time Session Counter
A SessionTokenCounter component is embedded in the chat interface, showing live token consumption for the current conversation.
Export
Export your usage data anytime:
- CSV — For spreadsheets and analysis
- JSON — For programmatic processing
Filter by time period before exporting.