OpenRouter Usage

Query OpenRouter API costs and generate usage reports via the command line.

When to Use

Checking OpenRouter spend over the last 30 days
Breaking down costs by model, provider, or date
Exporting usage data as CSV or JSON for accounting
Auditing which models and API keys consume the most credits

Prerequisites

OPENROUTER_API_KEY environment variable (or in a project .env file)
The key must be a Management key — create one at https://openrouter.ai/settings/management-keys
Regular provisioning keys return 401/403 on analytics endpoints

Usage

# Full 30-day summary with credits balance
python3 ~/.claude/skills/openrouter-usage/scripts/openrouter_usage.py

# Filter to a single date
python3 ~/.claude/skills/openrouter-usage/scripts/openrouter_usage.py --date 2026-02-01

# Export as CSV (for spreadsheets)
python3 ~/.claude/skills/openrouter-usage/scripts/openrouter_usage.py --csv

# Raw JSON output (for piping to jq)
python3 ~/.claude/skills/openrouter-usage/scripts/openrouter_usage.py --json

# Skip credits lookup (activity only)
python3 ~/.claude/skills/openrouter-usage/scripts/openrouter_usage.py --no-credits

Output

The default report includes:

Credits balance: total purchased, used, remaining
Spend by model: per-model cost with token counts (prompt, completion, reasoning)
Spend by provider: aggregated by provider (OpenAI, Anthropic, Google, etc.)
Spend by date: daily spend with visual bar chart

API Endpoints

Both endpoints are read-only GETs authenticated via Bearer token:

Endpoint	Purpose
`GET /api/v1/credits`	Total credits purchased and used
`GET /api/v1/activity`	Last 30 days of usage grouped by model/endpoint/provider

The activity endpoint returns per-row: date, model, provider_name, usage, byok_usage_inference, requests, prompt_tokens, completion_tokens, reasoning_tokens.