OpenRouter Cost Watch
Automated OpenRouter price monitoring and alerts with zero recurring token costs.
Free
Works with the AI tools you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+22 moreClaude CodeCursorCodex CLIMuseDots
OpenClaw
Manus
Grok
Grok Bot+18 agentsOpenRouter Cost Watch
Example session with this skill installed
Check the current OpenRouter prices for deepseek/deepseek-v3 and compare them against my baseline in ~/.openrouter-cost-watch/config.json.
- Read your context and instructions
- Compiled the openrouter cost watch
I have executed the price check against the OpenRouter API. The script compared live rates for your tracked models against the stored baseline and detected significant movement in one model. The baseline has been updated to reflect these new rates.
OpenRouter Price Alert
| Model ID | Direction | Type | Old Price ($/M) | New Price ($/M) | Change (%) |
|---|---|---|---|---|---|
deepseek/deepseek-v3 | ▲ | Input | $0.0140 | $0.0210 | +50.0% |
deepseek/deepseek-v3 | -- | Output | $0.0600 | $0.0600 | 0.0% |
qwen/qwen-2.5-72b | -- | Input | $0.4000 | $0.4000 | 0.0% |
qwen/qwen-2.5-72b | ▼ | Output | $0.4000 | $0.3200 | -20.0% |
Status: Alert triggered for deepseek/deepseek-v3 input price. The change of $0.0070/M exceeds your min_abs_usd floor of $0.0050.
Next steps
- Review the
deepseek/deepseek-v3price hike to ensure it still fits your application's budget. - Check if a lower-cost provider for the same model is available via OpenRouter's model list.
- No action needed for Qwen; the 20% drop is below your 25% alert threshold.
Connects securely to your tools. The creator never sees your data.
About this skill
What it does
OpenRouter Cost Watch tracks the per-million-token price of the LLM models you use and alerts you when that cost moves. It fetches OpenRouter's public model catalog, converts per-token rates to $/M, holds a rolling baseline, and only pings you when a tracked model's input or output price changes by more than your threshold.
Why it matters
LLM API providers reprice without fanfare. If you build on model APIs, your running cost can change under you and you'd never notice. This gives you a standing, hands-off watch so a repricing never catches you by surprise.
Key features
- Zero tokens to run - the recurring check is a plain stdlib Python script, no LLM in the loop
- Percentage + absolute-threshold alerting so cheap-model noise doesn't spam you
- Auto-updating baseline prevents duplicate alerts
- Sentinel prices (router aliases) automatically filtered
- No API key needed - OpenRouter's catalog is public
Setup
Copy config.example.json to your config, add the model IDs you want to track, run python3 watch.py init to snapshot a baseline, then schedule python3 watch.py on your preferred cadence (agent cron, system cron, etc.).
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
3 people have installed this skill.
Trust & safety
Security scanned
Verified clean 12 days ago
- Free to download with an account
Needs access to
Openrouter