Explain a run in plain language

Authoring & analysis

loadr explain reads a run's summary and reports the threshold verdict, error rate, latency tail, and a heuristic likely cause — saturation vs a slow path vs a healthy run.

Browse all 52 demos 12 categories

Load profiles

The 30-second quickstart
Ramping load
Open-model arrival rate
Spike test
Soak & outputs

Traffic modelling

Weighted traffic mix
Flow control & weighted tasks
Feeders & throttling
Data-driven load
Correlation & extraction

Validation & CI

Functional / smoke test
JMESPath & fused check-chains
Failure breakdown, by cause
Tags & per-route metrics

Authoring & analysis

Record a session into a scenario
Generate a test from an API contract
Explain a run in plain language
Catch regressions statistically

Auth, sessions & uploads

Per-VU auth tokens + refresh
Cookie-based sessions
Multipart file uploads
Advanced HTTP

Protocols

WebSockets
gRPC
GraphQL
Raw TCP & UDP
Server-Sent Events
SOAP / XML
Twirp RPC
Browser (headless)

Databases & messaging

PostgreSQL
MongoDB
MySQL
Redis
Apache Kafka
RabbitMQ
Elasticsearch

Scripting & metrics

JavaScript in the flow
Custom metrics
Lifecycle hooks

Scale & operations

Distributed fleet
Environments
Scenarios & groups
Time-series HTML report

Ops integrations

Prometheus + Grafana
CI/CD performance gate
Kubernetes

Adopt loadr

Import from k6 & JMeter
The live Web UI
The agent fleet, in the UI

Extending loadr

Write a plugin in Go
A native protocol in C

Run it

$ loadr explain summary.json

What it shows

▸Latency tail: p99 vs the median
▸Saturation vs slow-path heuristic
▸The offline half of the AI copilot

Download & run → Read the docs →