What it does
Blackfire combines lightweight application performance monitoring with deterministic and continuous profiling, performance tests, and recommendations. Its monitoring surfaces high-level performance signals from live application traffic; developers can then profile requests to inspect time spent in functions and external services. The product is designed for web application teams that want to move from detecting a performance regression to investigating the code behind it, and to keep performance budgets in CI workflows.
Blackfire links production monitoring to code-level profiling. Its monitoring data gives a high-level view of application performance and points toward transactions or services that consume resources. A deterministic profile can then expose function calls, caller-callee relationships, CPU and wall time, memory use, I/O, and external HTTP, SQL, or queue activity.
Move from a signal to a profile
The official installation guide describes a stack of components: a language probe or extension, an agent that communicates with Blackfire, and a profiling client such as the CLI or browser extension. Once those pieces are configured for an environment, enable monitoring to collect live performance metrics. If a slow request stands out, profile it and use the call graph or timeline to locate the work that accounts for its cost. Blackfire's documentation notes that monitoring is intentionally lightweight and less detailed than a profile, so the two views answer different questions.
Keep performance from slipping back
A team can add performance assertions to tests and connect them to CI/CD or deployment checks. For example, after finding a costly database call, an engineer could set an assertion around that route's performance budget and run it as part of the normal test pipeline. Blackfire also lists Synthetic User Monitoring as a QA option. These are documented uses, not test results from this research.
Blackfire's site focuses on PHP and Python applications. Its setup is not just a dashboard login: it requires appropriate probes and an agent, and the PHP installation notes warn about known conflicts with XHProf, Pinba, Suhosin, Uopz, Xdebug, and pcov. Pricing is subscription-based, with separate development and production plans; the pricing page also lists quotas and add-ons, so teams should check the current terms for their environment. Exact amounts are omitted here.
Strengths
- Combines low-overhead live performance monitoring with a profiler that can trace a problem down to functions and external service calls.
- Supports performance assertions, testing, and CI/CD integration so a team can check budgets before or after deployment.
- Provides both continuous profiling and synthetic monitoring as part of a broader application performance workflow.
Watch out for
- Monitoring is less detailed than deterministic profiling, so deeper code-level diagnosis requires a profile in addition to the high-level signal.
- Setup involves language-specific probes or extensions, an agent, and a CLI or browser client; the PHP installation guide documents extension conflicts that teams should check against their runtime.
Best for
- PHP or Python teams investigating slow web requests in production.
- Developers who want to compare profiling details with live performance signals.
- Teams adding performance assertions and synthetic checks to a CI/CD workflow.
Sources
- Crush performance bottlenecks | Blackfire PHP & Python Observability (Oct 26, 2026)
- Application Performance Monitoring (Oct 26, 2026)
- What is Blackfire? (Oct 26, 2026)
- Installing Blackfire (Oct 26, 2026)
- Pricing (Oct 26, 2026)