QuantuisBenignus

• Actual Electricity Consumption and Cost to Run Local LLMs. From Gemma3 to QwQ.

  Posted by QuantuisBenignus@reddit | LocalLLaMA | View on Reddit | 52 comments

• Local speech chat with Gemma3, speaking like a polyglot with multiple-personalities

  Posted by QuantuisBenignus@reddit | LocalLLaMA | View on Reddit | 0 comments

• Tokens per WattHour (Tokens per cent) for various local LLMs

  Posted by QuantuisBenignus@reddit | LocalLLaMA | View on Reddit | 3 comments

• Practical tip: llama.cpp aliases for LLMs that require direct system prompt at the command line

  Posted by QuantuisBenignus@reddit | LocalLLaMA | View on Reddit | 2 comments

• Speaking with your local LLM with a key press on Linux. Speech to text into any window, most desktop environments. Two hotkeys are the UI.

  Posted by QuantuisBenignus@reddit | LocalLLaMA | View on Reddit | 3 comments