Lanbench -

async def main(): # Initialize LANBench with all features dashboard = LiveDashboard() api_server = FastAPI() distributed_tester = DistributedTester()

is a free, lightweight network benchmarking utility designed to test the TCP performance LANBench

Keywords used: LANBench, local AI benchmarking, LLM performance testing, OpenAI compatible API, tokens per second, time to first token, llama.cpp benchmark, Ollama test, AI latency measurement. async def main(): # Initialize LANBench with all

return 'idle_latency': idle_latency, 'loaded_latency': loaded_latency, 'bufferbloat_ms': loaded_latency - idle_latency Look for a stream: true variable

LANBench expects the API to return a stream of Server-Sent Events (SSE), but your server returns a single JSON blob. Solution: Check your LANBench configuration. Look for a stream: true variable. If your server doesn't support streaming, set it to false .

You open LANBench via file:// and try to hit http://localhost:8080 . The browser blocks it. Solution: