Le Répertoire des AgentsFR · EN · ES · JA · HI
Home › Local AI › Advanced agents

“I want to chat with an AI without my texts leaving for the internet”

Anyone with a recent computer who wants to try a 100% private AI, with no account or subscription — the simplest starting point for a local AI.

Recommended agent: LM Studio
💰 $0: the application is free, including for use at work (confirmed on the publisher's official pages). Level: Easy 🌍 United States Free with no time limit. Proprietary application (not open source); paid cloud options exist but are unnecessary for local use

Price verified on October 31, 2026.

LM Studio (lmstudio.ai), Element Labs' reference application for running open models (Llama, Qwen, DeepSeek, Gemma…) directly on your computer: a built-in model catalog, a familiar chat interface, and the ability to query your documents offline. Everything happens on your machine.

How it works

Trigger: You open the app and ask your question in the chat window — loads the chosen model into your computer's memory and generates the answer entirely locally, and you get answers generated at home, with no account, no subscription and no data sent anywhere.

Step-by-step setup

  1. Download the official installer for your system (lmstudio.ai) and launch the application.
  2. In the search tab, download a first model suited to your machine — start small: 4 to 8 billion parameters.
  3. Open a chat and ask your questions: everything happens on your machine.
  4. Drop a document into the chat to query it offline.
⚠️ Guardrails: Intel Macs aren't supported, and an old computer will make for a disappointing experience: first check that you have an Apple Silicon Mac or a recent PC with 16 GB of RAM — otherwise, start with a very small model.
Find your agent with the quiz →

Frequently asked questions

Is LM Studio really free, even at work?

Yes: the vendor states that LM Studio is free, including for professional use. You only pay if you plug in paid online models — which isn't necessary for local use.

Which first model should I download?

A model with 4 to 8 billion parameters (4B/8B) in quantized version: about 3 to 6 GB, the right balance on a machine with 16 GB of RAM. Our “Installing an AI on Your Computer” guide details the choice.

My Mac is an Intel, not Apple Silicon?

LM Studio targets Apple Silicon Macs and recent PCs. On an Intel Mac, turn instead to Jan or GPT4All, which are more forgiving about hardware.

Other agents for this problem

LLM Council

I want a solid second opinion before a big decision

Free software (public repository, no open-source license); you pay for the model calls (about 2 to 3 times the cost of asking a single model).

CrewAI

I want a team of agents that splits up a big job

$0 for the open-source framework and the basic platform (50 runs/month); enterprise platform by quote, plus the cost of the model calls.

LangGraph

I want a custom agent workflow, with every step under control

Free open-source library; you pay for the model calls, and for LangSmith if you take it (free to start, then $39/user/month).

AutoGen

I want agents that debate each other until they find the solution

Free and open source; you pay for the model calls used by the agents — a multi-round debate costs several times a simple answer.

MCP (Model Context Protocol)

I want to plug my own tools and data into my agents

Free: it's an open standard, not a product; each tool you plug in keeps its own pricing.

See also

“I want a solid second opinion before a big decision” — LLM Council“I want a team of agents that splits up a big job” — CrewAI“I want a custom agent workflow, with every step under control” — LangGraph

Profile updated on November 1, 2026 — see what's new at the Directory.