Skip to content
All Tutorials

Connect a Local Ollama Model to ClawNest

Install Ollama locally and connect it to a new AI assistant in dash.clawnest.ai using this practical step-by-step guide.

What you'll need

  • A computer with at least 32 GB RAM or a GPU with 24 GB+ VRAM to run a local model
  • A free Tailscale account — tailscale.com
  • A ClawNest account — dash.clawnest.ai (free Starter plan works)
  • About 15 minutes to complete the setup

Part 1: Install Ollama Locally

Install Ollama and Tailscale on the machine that will run your local model.

  1. 1

    Install Ollama from the official download page

    Go to ollama.com/download, choose your operating system, and complete the installer.

  2. 2

    Start Ollama with network access enabled

    Run OLLAMA_HOST=0.0.0.0 ollama serve so it listens on all network interfaces, not just localhost (port 11434). Setup varies per OS — see the Ollama FAQ for details.

  3. 3

    Download a model that can handle agent workflows

    Run ollama pull qwen3:32b as a safe default. You can select another supported model later.

  4. 4

    Install Tailscale on the Ollama machine

    Download Tailscale from tailscale.com/download, install it, and log in. This creates a secure private network between your devices.

  5. 5

    Find your Tailscale IP address

    Go to login.tailscale.com/admin/machines and note the IP address (e.g. 100.75.36.94) of the machine running Ollama. You will need this in Part 2.

Example commands

ollama pull qwen3:32bOLLAMA_HOST=0.0.0.0 ollama serve
Tailscale admin console showing machines and their Tailscale IP addresses.
Find your Ollama machine's Tailscale IP under Machines at login.tailscale.com/admin/machines.

Official downloads: ollama.com/download  |  tailscale.com/download

Part 2: Create an Assistant in ClawNest Dashboard

Once Ollama is ready, log in and connect your local model when creating the assistant.

ClawNest dashboard assistant form with BYOK billing and Ollama provider selected.
Steps 1-2 in the dashboard: select BYOK and then Ollama.
ClawNest dashboard assistant form with Ollama server URL, model selection, and create button.
Steps 3-5 in the dashboard: URL, model, then create assistant.
  1. 1

    Set Billing Mode to Bring Your Own Key

    In the assistant form, choose "Bring Your Own Key".

  2. 2

    Choose Ollama as AI Provider

    In the provider section, click "Ollama".

  3. 3

    Add your Ollama Server URL

    Enter your URL, for example http://100.x.x.x:11434 (your Tailscale IP or hostname). /v1 is appended automatically.

  4. 4

    Select the model

    Choose the same model you downloaded in Ollama, for example `Qwen 3`.

  5. 5

    Click Create Assistant

    Review the form and click "Create Assistant" to finish setup.

Log in first at dash.clawnest.ai

Part 3: Connect Your Assistant to Tailscale

After creating the assistant, connect it to your Tailscale network so it can reach your local Ollama instance.

ClawNest dashboard assistant card with the Connect button highlighted.
Step 6: Click Connect on your assistant.
Tailscale Network dialog showing Not Connected status with Connect to Tailscale button.
Step 7: Click Connect to Tailscale.
Connect to Your Tailnet dialog showing instructions and Open Tailscale Console link.
Step 8: Click Open Tailscale Console to generate an auth key.
Tailscale console showing Generate auth key dialog with description field and Generate key button.
Steps 9-10: Generate an auth key in the Tailscale console.
Tailscale dialog showing the generated auth key ready to copy.
Copy the generated auth key (tskey-auth-...).
ClawNest dialog with the auth key pasted and Connect to Tailscale button.
Step 11: Paste the key and click Connect to Tailscale.
Tailscale Network dialog showing Connected status with hostname, IP address, and tailnet.
Step 12: Connection verified — your assistant is on the Tailscale network.
ClawNest dashboard showing assistant online on the tailnet with Open Chat button.
Step 13: Your assistant is online. Click Open Chat to start.
  1. 6

    Click Connect on the assistant

    On the assistant card in the dashboard, click the "Connect" button.

  2. 7

    Click Connect to Tailscale

    In the Tailscale Network dialog, click "Connect to Tailscale".

  3. 8

    Open Tailscale Console to get an auth key

    Click "Open Tailscale Console" to open the Tailscale admin panel where you generate an auth key.

  4. 9

    Generate an auth key in Tailscale

    In the Tailscale console, click "Generate auth key...", give it a description, then click "Generate key". Copy the generated key (starts with tskey-auth-).

  5. 10

    Copy the generated auth key

    Copy the key from the dialog. It will only be shown once.

  6. 11

    Paste the auth key and connect

    Back in ClawNest, paste the auth key into the field and click "Connect to Tailscale".

  7. 12

    Verify the connection

    The dialog now shows "Connected" with your hostname, IP address, and tailnet. Your assistant is online on your Tailscale network.

  8. 13

    Open Chat and start using your local model

    Back on the assistant card you'll see it's online on your tailnet. Click "Open Chat" to start chatting with your local Ollama model.

OpenClaw Gateway chat interface ready to send a message to the local Ollama model.
You're connected! Start chatting with your local Ollama model.
Start Free Trial

Next Steps

Now your local model is connected. Explore use cases or compare plans for production workloads.

All TutorialsView Pricing