If you've ever been surprised by an AI API bill at the end of the month, TokenWatch is built for exactly that problem. The new tool connects to your OpenAI, Anthropic, and Google Gemini API keys and provides a single dashboard showing real-time spend across all three.
The feature list covers the basics you'd expect from a cost monitor, plus a couple of useful additions:
- Real-time spend dashboard across all connected projects
- Budget thresholds with proactive alerts before you hit limits
- Anomaly detection that flags unusual cost spikes
- Duplicate call detection for catching redundant API requests
That last one is particularly practical. Duplicate API calls are one of the most common sources of wasted spend in AI applications, especially during development when retry logic or poorly cached prompts can silently double or triple your costs.
The free tier supports monitoring up to three AI provider connections. The tool was built as a solo project in roughly two days, so expect a relatively lean feature set at launch. But for small teams and individual developers running API-heavy workloads, even basic visibility into spend-per-project is a meaningful upgrade over checking your provider dashboards separately.
The AI API cost monitoring space is getting more crowded as companies realize their inference costs are a real line item now. Tools like Helicone and LangSmith offer similar observability features, though they tend to focus more on prompt engineering and evaluation alongside cost tracking. TokenWatch appears to be positioning itself as a simpler, cost-focused alternative.