Every request.
Fully visible.
Logs, alerts, and callbacks for every LLM call. See what happened, get notified when things break, and pipe data to your existing tools.
Observability stack
Logs. Alerts. Callbacks. Control.
Request Logs
searchable · filterable · expandableEvery LLM request logged with model, status, latency, cost, tokens, and full prompt/response (if data policy allows). Search by request ID, model, status, or time range.
| Request | Model | Status | Latency | Cost |
|---|---|---|---|---|
| req_a1b2 | gpt-4o | 200 | 1.2s | $0.032 |
| req_c3d4 | claude-3.5-sonnet | 200 | 0.8s | $0.018 |
| req_e5f6 | gpt-4o | 400 | 0.1s | $0.000 |
| req_g7h8 | gemini-1.5-pro | 200 | 2.1s | $0.041 |
Logging Callbacks
connect · forwardPipe logs to your existing observability stack. One toggle per provider.
8 Alert Types
toggle · per-orgLLM errors, budget warnings, provider outages, and more. Toggle each independently.
Data Policy Control
per-org · complianceChoose exactly what gets logged. Full prompts and responses, metadata only, or nothing at all. Compliance-friendly, SOC2-ready.
Prompts + responses + metadata
Model, tokens, cost, latency
Nothing stored
Notification Channels
multi-channelGet alerts where your team lives. Email, Slack, Teams, or custom webhooks.
Real-time
sub-second · searchableLogs appear within milliseconds. Search, filter, and drill into any request as it happens.
See everything. Miss nothing.
Full observability on every plan. Logs, alerts, callbacks, and data policy controls included.