# Atlarix Auto: A Free AI Coding Agent That Picks the Model for Each Step > Source: > Published: 2026-09-23 20:57:54+00:00 Most AI coding tools ask you to make two decisions before you write a line: which model, and whose API key. Atlarix Auto removes both. It is one option in the model picker — the one new installs start on — and instead of running your whole session on a single model, it chooses a model for each step of the work. On an account with no credit, Auto in the Atlarix desktop app is free. On Atlarix Pro or credit, it is paid and private. Atlarix is a desktop coding agent: it plans, edits your code, runs commands in a per-OS sandbox, and runs your tests before it calls anything done, with every change waiting for your approval. Auto is the model setting that decides which language model does each piece of that work. You pick Auto and give it a task. A coding session is not one kind of work. Reading a codebase, writing a change, planning a refactor and reviewing a diff need very different amounts of reasoning. Auto routes on what each request is: | What the step is | What Auto uses | |---|---| | Exploring and building | A fast, affordable model that is strong at code | | Planning and code review | A stronger reasoning model | | Deep thinking switched on | Every step moves up one level | | A step with a screenshot, image or PDF | A model that can actually read it | | Sub-agents, session titles, summaries | The cheapest capable model | | A model is rate-limited | The same level from another provider | Routing happens per request, not per session: a long task that suddenly needs to look at a screenshot gets a vision-capable model for that one step, then returns to the cheaper one. If a step fails, Atlarix retries it; a "Try with a stronger model" button reruns a hard turn one level up. On an account with no credit, Auto in the desktop app costs nothing. It runs on OpenAI models through OpenAI's data-sharing program: OpenAI provides free daily capacity for traffic it may use to improve its models. That is the whole reason it is free. Before your first free message, Atlarix shows a one-time notice explaining this, and you choose whether to go ahead. Free Auto is available in the desktop app (and to Atlarix Reviewer for public repositories only); the Chrome extension runs on paid Auto. Each account gets a fair daily share, sized from what is left and how many people are using it, with a floor so everyone gets real work done and a ceiling so no single account drains it. When your share runs out, Auto falls back to a free model rather than cutting you off; only past that does it pause until the daily reset — shown on your own clock. With Atlarix Pro or credit, Auto is paid and private: nothing it handles is shared or used to train models. Your own API keys work as before. Local models through Ollama or LM Studio never leave your machine. Atlarix itself never trains on your code, in any mode. The app is free — every feature, no gating. Paid inference is Atlarix Pro ($19/month, including Core usage across the desktop app, the Chrome extension and Atlarix Reviewer) or pay-as-you-go credit, with the cost of every reply shown. Download Atlarix from [atlarix.dev](https://www.atlarix.dev). New installs start on Auto; if you picked a model before, switch to Auto in the model picker, then give it a real task on your own code. *Originally published at [atlarix.dev](https://www.atlarix.dev/blogs/atlarix-auto-free-ai-coding-agent-no-api-key).*