bench(python): use pyperf strict mode over custom logic

Author: 0x676e67

README.md

@@ -107,7 +107,7 @@ maturin build --release
 
 ## Benchmark
 
-Outperforms `requests`, `httpx`, `aiohttp`, and `curl_cffi`, and you can see the [benchmark](https://github.com/0x676e67/wreq-python/tree/main/bench) for details — benchmark data is for reference only and actual performance may vary based on your environment and use case.
+Outperforms `requests`, `httpx`, `aiohttp`, and `curl_cffi`, and you can see the [benchmark](https://github.com/0x676e67/wreq-python/tree/main/bench) for details. The benchmark suite is driven by `pytest-benchmark`, and results are still for reference only because network and system conditions affect outcomes.
 
 ## Services
 

bench/README.md

@@ -40,7 +40,6 @@ Sync clients
 - requests
 - niquests
 - pycurl
-- [python-tls-client](https://github.com/FlorianREGAZ/Python-Tls-Client.git)
 - httpx
 - wreq
 - ry
@@ -64,12 +63,12 @@ All the clients run with session/client enabled.
 ## Run benchmark
 
 ```bash
-# Install dependencies  
-pip install -r requirements.txt
+# Install project + benchmark dependencies
+pip install -e .[bench]
 
 # Start server
 python server.py
 
-# Start benchmark
+# Run benchmark suite
 python benchmark.py
-```
+```