Usage & Limits
What counts as an event, how monthly quotas work, and what happens when you go over.
What counts as an event
Every piece of data you send to DeepTracer is one event. Specifically:
| Data type | What counts as 1 event |
|---|---|
| Log entry | Each individual log in a batch |
| Error report | Each error sent to /ingest/errors |
| Trace span | Each span sent to /ingest/traces |
| LLM usage report | Each LLM call sent to /ingest/llm |
Batching and event counts
The log endpoint accepts batches of up to 1,000 entries per request. Each entry in the batch counts as a separate event.
- Sending 1 request with 100 logs = 100 events (1 HTTP request)
- Sending 100 requests with 1 log each = 100 events (100 HTTP requests)
Both use the same number of events. Batching reduces HTTP overhead, not event counts.
Monthly event limits
| Plan | Events per month |
|---|---|
| Free | 25,000 total |
| Pro | 2,000,000 per project |
Events reset on the first day of each billing cycle.
What happens when you hit the limit
Free plan — hard cutoff
When you reach 25,000 events, the SDK stops sending new data for the rest of the month.
- Your app keeps running normally — the SDK never crashes or slows down your code
- The dashboard still shows all data collected before the cutoff
- Existing AI investigations and chat history remain accessible
- Events resume automatically on the next billing cycle
Pro plan — soft overages
DeepTracer never cuts off your data on the Pro plan. Monitoring is too important to stop because of a billing counter.
If you go over 2,000,000 events in a month:
- Data continues flowing with no interruption
- Excess events are billed at $0.50 per 100,000 events
- Overage charges appear on your next invoice
For context: 2M events covers most apps comfortably. A typical Next.js app sending logs, errors, and traces generates roughly 500K-1M events per month.
AI limits
Investigations
| Plan | Investigations per month |
|---|---|
| Free | 3 |
| Pro | Unlimited |
An investigation is triggered when you click "Investigate" on an error group, or when the AI agent automatically investigates a critical error (Pro only).
Chat messages
| Plan | Messages per month |
|---|---|
| Free | 10 |
| Pro | Unlimited |
Each message you send in the AI chat counts toward your limit. The AI's responses don't count.
Data retention
| Data type | Free | Pro |
|---|---|---|
| Logs | 1 day | 7 days |
| Errors | 1 day | 30 days |
| Traces | 1 day | 7 days |
| LLM usage | 1 day | 7 days |
| Investigations | 1 day | 30 days |
After the retention period, data is permanently deleted. On Pro, errors are kept for 30 days (longer than other data types) because error patterns are valuable for longer-term debugging.
Other limits
| Resource | Free | Pro |
|---|---|---|
| Health check endpoints | 5 | 50 |
| Team seats | 1 | 5 |
| Projects | 1 | Unlimited |
| API key requests | 5,000/min | 5,000/min |
| Max request body | 5 MB | 5 MB |
| Max logs per batch | 1,000 | 1,000 |
Checking your usage
You can see your current event count and AI usage in the dashboard:
- Open Settings in the sidebar
- Your current billing period usage is shown at the top
The event counter updates in near real-time.