I gave my local LLM a system prompt modeled on how Claude talks, and it stopped feeling like a cheap substitute
Getting my 4B and 9B models to talk like Claude
When you connect a local LLM to your machine, it can feel less personal than using a cloud-based model. Local models require more effort from the user, and their output can be inconsistent. However, by customizing the system prompt, users can make their local models behave more like popular cloud models such as Claude. Claude's unique personality, which includes being knowledgeable, direct, and honest, can be replicated through specific system prompts.
These prompts dictate how the model should respond, such as by answering first and pointing out flaws instead of agreeing. By tailoring the system prompts for different tasks, users can optimize their local models to sound like Claude, whether for coding-related work or general conversation.
Written by urgent.news from XDA Developers's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
