The latest articles on DEV Community by Escape AI (@escape_ai_a7b4027bc8015a2). https://dev.to/escape_ai_a7b4027bc8015a2 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4113610%2Febf95ab4-3a72-4902-821e-c65c2e7dfb9d.png https://dev.to/escape_ai_a7b4027bc8015a2 en Escape AI Mon, 07 Sep 2026 09:35:59 +0000 https://dev.to/escape_ai_a7b4027bc8015a2/free-local-ai-coding-with-ollama-what-actually-runs-on-your-machine-59l8 https://dev.to/escape_ai_a7b4027bc8015a2/free-local-ai-coding-with-ollama-what-actually-runs-on-your-machine-59l8 <p>Escape AI can connect to a local <a href="https://ollama.com" rel="noopener noreferrer">Ollama</a> install so Ask, Plan, and text-based Edit run against a model on your own hardware instead of a cloud API — no per-request cost, and nothing sent off your machine for those requests. Here's exactly what that covers, what still uses a cloud model, and how to turn it on.</p> <h2> What "local" actually means here </h2> <p>Local means the model itself runs on hardware you control — your laptop, workstation, or a server you operate — via Ollama. Escape AI never assumes a fixed model name: it asks your Ollama install which models are actually pulled and lists exactly those, so the picker always reflects what you really have installed, not a hardcoded suggestion.</p> <h2> What runs locally, and what doesn't </h2> <p>Ask mode and Plan mode both work fully against a local model — plain chat and structured multi-step plans don't need function-calling, so a local model handles them the same way a cloud model would. Text-based Edit (asking for a complete rewritten file, without tool calls) also works locally.</p> <p>Agent mode — the real tool-calling loop that reads files, applies diffs, and runs terminal commands — stays on a cloud model. Reliable function-calling (the model correctly calling a tool instead of just describing what it would do) is inconsistent enough across local and non-OpenAI models that this app doesn't route that specific loop through anything but a proven tool-calling model, regardless of which model you've picked elsewhere in the composer.</p> <h2> Setting it up </h2> <ol> <li> <strong>Install Ollama and pull a model.</strong> Install from <a href="https://ollama.com" rel="noopener noreferrer">ollama.com</a>, then pull a coding-capable model, e.g. <code>ollama pull qwen2.5-coder</code>. Any model you've pulled shows up automatically in the picker.</li> <li> <strong>Point Escape AI at it.</strong> Defaults to <code>http://localhost:11434</code> (Ollama's own default).</li> <li> <strong>Pick it from the model picker.</strong> A "Local" section lists your real, discovered Ollama models.</li> </ol> <h2> When to reach for a cloud model instead </h2> <ul> <li>Agent mode's real tool-calling loop — always cloud, by design.</li> <li>When your local hardware can't comfortably run a large enough model for the task.</li> <li>When you're already paying for a cloud plan and prefer its speed/quality trade-off for everyday use.</li> </ul> <p>Try it: <a href="https://escape-ai.org" rel="noopener noreferrer">escape-ai.org</a> — free to start, no card required.</p> ai productivity webdev ollama