Llama 3.1 70B: Local Ollama Speed Boost

Is it just me or is running Llama 3.1 70B locally via Ollama incredibly time-efficient – I’m seeing a 90% reduction in API call latency for testing?