💰 cost-tracker

Track AI API costs across providers — Anthropic, OpenAI, OpenRouter. Daily/weekly/monthly spend reports.

🦞 Pixeyo CuratedRequires: curl✓ Read source below

Track and report AI API spending across multiple providers. Helps users stay within budget and catch unexpected cost spikes.

When to Use

Use when the user asks about API costs, spending, usage, billing, budget, or wants to check how much they've spent on AI providers.

Supported Providers

Anthropic

curl -s https://api.anthropic.com/v1/usage \
  -H "x-api-key: ${ANTHROPIC_API_KEY}" \
  -H "anthropic-version: 2023-06-01" \
  --data '{"start_date":"'"$(date -u -v-30d +%Y-%m-%d 2>/dev/null || date -u -d '30 days ago' +%Y-%m-%d)"'","end_date":"'"$(date -u +%Y-%m-%d)"'"}'

If the API doesn't support this endpoint, direct the user to:

https://console.anthropic.com/settings/usage

OpenAI

open "https://platform.openai.com/usage"

Direct the user to check their dashboard at https://platform.openai.com/usage.

OpenRouter

curl -s https://openrouter.ai/api/v1/auth/key \
  -H "Authorization: Bearer ${OPENROUTER_API_KEY}" | python3 -m json.tool

This returns usage data including total credits used. Dashboard: https://openrouter.ai/activity

Manual Budget Tracking

If API endpoints aren't available, help the user maintain a simple cost log:

echo "$(date +%Y-%m-%d),anthropic,$(echo 'estimated_cost'),note" >> ~/.openclaw/workspace/cost-log.csv

Cost estimation by model (approximate per 1M tokens):

Model	Input	Output
claude-sonnet-4-6	$3	$15
claude-opus-4-6	$15	$75
gpt-5.2	$2.50	$10
gpt-4.1	$2	$8
gemini-3.1-pro	$1.25	$5

Response Format

Present spending as a clean summary:

💰 AI Spending Report (Last 30 days)
━━━━━━━━━━━━━━━━━━━━━━
Anthropic:    $32.40 (Sonnet 4.6 primary)
OpenRouter:   $8.15  (fallback usage)
Total:        $40.55
Budget:       $50/month
Remaining:    $9.45 (19%)
━━━━━━━━━━━━━━━━━━━━━━
💡 Tip: Heartbeats account for ~$4/mo. Use a cheaper model.

Rules

Never expose or log full API keys — only use environment variables.
If a provider API doesn't return cost data, direct the user to the web dashboard.
Proactively warn if spend exceeds 80% of a stated budget.
Suggest cost optimizations: cheaper heartbeat models, lower fallback frequency, cache-ttl mode.

📥 Download All Skills (.zip)🧩 Skill Explorer

📄 View raw SKILL.md source

--- name: cost-tracker description: "Track AI API costs across providers — Anthropic, OpenAI, OpenRouter. Daily/weekly/monthly spend reports." user-invocable: true metadata: openclaw: requires: bins: ["curl"] author: "pixeyo" homepage: "https://openclawcheatsheet.com/skills" --- # cost-tracker Track and report AI API spending across multiple providers. Helps users stay within budget and catch unexpected cost spikes. ## When to Use Use when the user asks about API costs, spending, usage, billing, budget, or wants to check how much they've spent on AI providers. ## Supported Providers ### Anthropic ```bash curl -s https://api.anthropic.com/v1/usage \ -H "x-api-key: ${ANTHROPIC_API_KEY}" \ -H "anthropic-version: 2023-06-01" \ --data '{"start_date":"...","end_date":"..."}' ``` If the API doesn't support this endpoint, direct the user to: - https://console.anthropic.com/settings/usage ### OpenAI ```bash open "https://platform.openai.com/usage" ``` ### OpenRouter ```bash curl -s https://openrouter.ai/api/v1/auth/key \ -H "Authorization: Bearer ${OPENROUTER_API_KEY}" | python3 -m json.tool ``` ## Cost estimation by model (approximate per 1M tokens): | Model | Input | Output | |-------|-------|--------| | claude-sonnet-4-6 | $3 | $15 | | claude-opus-4-6 | $15 | $75 | | gpt-5.2 | $2.50 | $10 | | gpt-4.1 | $2 | $8 | | gemini-3.1-pro | $1.25 | $5 | ## Response Format ``` 💰 AI Spending Report (Last 30 days) ━━━━━━━━━━━━━━━━━━━━━━ Anthropic: $32.40 (Sonnet 4.6 primary) OpenRouter: $8.15 (fallback usage) Total: $40.55 Budget: $50/month Remaining: $9.45 (19%) ━━━━━━━━━━━━━━━━━━━━━━ 💡 Tip: Heartbeats account for ~$4/mo. Use a cheaper model. ``` ## Rules - Never expose or log full API keys — only use environment variables. - If a provider API doesn't return cost data, direct the user to the web dashboard. - Proactively warn if spend exceeds 80% of a stated budget. - Suggest cost optimizations: cheaper heartbeat models, lower fallback frequency, cache-ttl mode.