Blackfire Profiler

Measure & understand code performance & behaviour in dev, test/staging & production.

Supported languages

PHP
Python

Code consumes server resources. Blackfire tells you how.

Blackfire Profiler is a tool that measures how your code consumes resources at run-time. It enables to find performance bottlenecks and understand the code's behavior.

Intuitive visualizations enable you to browse through your call stack, and review Wall-time, CPU time, I/O time, Memory, Network calls, HTTP requests, and SQL queries.

Not your usual APM.

Application Performance Monitoring tools are quite common nowadays. They are very different from Profilers - and they are complementary.

APMs gather performance metrics from real traffic on production. To avoid impacting the real end-users (with an "overhead"), they cannot gather too many back-end code metrics.

Blackfire Profiler's automatic instrumentation technology avoids any overhead on end-user traffic. Better, it is built for developers to measure and improve code anywhere it is deployed.

Build faster code by measuring its performance across iterations in development. Validate code changes in test/staging. Fix issues by measuring in production, where it all happens.

Automated issue detection

Tight integrations with languages and frameworks enable Blackfire to automatically detect issues in the code and offer resolution recommendations.

Each Blackfire Recommendation is fully documented, can be customized or disabled depending on the developer's needs.

Continuous Performance Optimization

On-demand profiling is great for ironing out a specific issue. But performance should be seen as a feature. It should be tested as such.

Blackfire offers a wide range of profiling automations and integrations, so that across deployments in production, or code iterations in testing, you can verify if your application performs as expected.

Get the most out of automations with Blackfire's detection of performance, security and quality issues, and related optimization recommendations.