Rust vs Competitors in 2025: Benchmark Results

Benchmark 2025: Rust vs C++, Go and Python — which is faster and more convenient?

In 2025, the ecosystem of systems programming languages continues to change. Rust, once a niche language, is increasingly chosen for high-load projects. We analyzed the latest benchmark results published in TechEmpower Round 25 and independent tests from The Computer Language Benchmarks Game to understand how much Rust outperforms its competitors and where it still falls short.

Testing Methodology

All tests were conducted on identical hardware: Intel Xeon Platinum 8480C (Sapphire Rapids), 64 GB RAM, Ubuntu 24.04 LTS. Measured metrics included: throughput (requests/sec), latency, memory consumption, and compilation speed. Typical scenarios were considered: JSON serialization, database operations, cryptography, and HTML parsing.

Results: Performance

• JSON serialization (simd-json): Rust — 1,850,000 req/s, C++ (nlohmann) — 1,720,000 req/s, Go — 980,000 req/s, Python (orjson) — 210,000 req/s. Rust is 7.5% faster than C++.
• PostgreSQL queries (50 connections): Rust (sqlx) — 48,000 tps, Go (pgx) — 41,000 tps, C++ (libpq) — 44,000 tps. Rust leads due to efficient memory management.
• Memory consumption (typical web application): Rust — 12 MB, Go — 22 MB, C++ — 14 MB, Python — 98 MB. Rust is nearly twice as memory-efficient as Go.

Where does Rust fall short?

Despite winning in raw performance, there are areas where competitors are stronger:

• Compilation speed: Rust compiles on average 3 times slower than C++ (with a full rebuild) and 10 times slower than Go. For fast iterations, Go remains king.
• Ease of prototyping: Python allows implementing the same task 5-7 times faster in terms of development time. Rust requires thoughtful design.
• Library ecosystem: C++ (Boost, Qt) and Go (built-in stdlib) have more mature solutions for GUI and embedded systems.

Conclusions for Python developers

If you use Python for backend development and are considering improving performance, Rust is the best candidate for replacing critical sections via PyO3 or maturin. It provides a 5-20x speedup on data operations and cryptography. However, for quick prototypes and microservices with frequent changes, Go remains more practical.

In 2025, Rust has firmly established itself in the top three for performance, second only to assembly and optimized C. For tasks where every millisecond and megabyte matters, Rust is the go-to choice.

Sources: TechEmpower Round 25 (2025), The Computer Language Benchmarks Game, own measurements on PythonLib.ru.