docs/guides/instinct.mdx

Rama main · réplica de lectura del repositorio original.

---
title: "Using Instinct with Ollama in Continue"
description: "Learn how to run Instinct, Continue's leading open Next Edit model, on your own hardware with Ollama"
---

<Warning>
  Instinct is a 7 billion parameter model. You should expect slow responses if
  running on a laptop. To learn how to inference Instinct on a GPU, see our
  [HuggingFace model card](https://huggingface.co/continuedev/instinct).
</Warning>

We recently released Instinct, a state-of-the-art open Next Edit model. Robustly fine-tuned from Qwen2.5-Coder-7B, Instinct intelligently predicts your next move to keep you in flow. To learn more about the model, check out [our blog post](https://blog.continue.dev/instinct/).

<Frame>
  <img src="/images/instinct.gif" />
</Frame>

### 1. Install Ollama

If you haven't already installed Ollama, see our guide [here](./ollama-guide).

### 2. Download Instinct

```bash
ollama run nate/instinct
```

### 3. Update your `config.yaml`

Open your `config.yaml` and add Instinct to the models section:

```yaml
# ... rest of config.yaml ...

models:
  - uses: continuedev/instinct
```

Alternatively, you can just click to add the block at https://continue.dev/continuedev/instinct.