Skip to main content

Using Ollama with YesBill

Ollama lets you run AI models locally on your own machine — completely free, no API keys, no usage limits. YesBill connects to your local Ollama instance through a secure tunnel so the cloud backend can reach it.

Step 1 — Install Ollama​

  1. Go to https://ollama.com/download
  2. Download the Windows installer and run it
  3. Ollama installs silently and adds a tray icon

Verify installation:

ollama --version

Step 2 — Pull a Model​

Download a model before starting the server. Some good options:

ModelCommandSizeBest for
Gemma 4ollama pull gemma4:latest~5 GBGeneral use
Llama 3.2ollama pull llama3.2~2 GBFast responses
Qwen3 Coderollama pull qwen3-coder:latest~4 GBCode & analysis
Gemma 3ollama pull gemma3:latest~3 GBBalanced
ollama pull gemma4:latest

List your installed models anytime:

ollama list

Step 3 — Start Ollama with Tunnel Support​

The Ollama tray app binds to 127.0.0.1 only, which blocks external tunnel traffic. You must start it manually with 0.0.0.0 binding.

First, quit the Ollama tray app (right-click the tray icon → Quit, or run):

taskkill /F /IM "ollama app.exe" /T
taskkill /F /IM ollama.exe /T

Then start Ollama with the correct settings:

set OLLAMA_HOST=0.0.0.0:11434 && set OLLAMA_ORIGINS=* && "C:\Users\%USERNAME%\AppData\Local\Programs\Ollama\ollama.exe" serve

You should see in the output:

OLLAMA_HOST:http://0.0.0.0:11434
Listening on [::]:11434

Keep this window open — Ollama must stay running while you use YesBill.

Step 4 — Install Cloudflare Tunnel​

YesBill's backend is hosted in the cloud. It cannot reach localhost on your machine. Cloudflare Tunnel creates a secure public URL that forwards traffic to your local Ollama.

Install via winget:

winget install Cloudflare.cloudflared

If cloudflared isn't recognized after install, use the full path:

C:\Program Files (x86)\cloudflared\cloudflared.exe

Add to PATH permanently (run once in PowerShell):

[Environment]::SetEnvironmentVariable("Path", $env:Path + ";C:\Program Files (x86)\cloudflared", "User")

Restart any terminal after this.

Step 5 — Start the Tunnel​

In a new CMD or PowerShell window (keep the Ollama window open):

cloudflared tunnel --url http://localhost:11434

Or with full path if not in PATH:

"C:\Program Files (x86)\cloudflared\cloudflared.exe" tunnel --url http://localhost:11434

After a few seconds you'll see:

Your quick Tunnel has been created! Visit it at:
https://some-words-here.trycloudflare.com

Copy that URL — you'll need it in the next step.

caution

The tunnel URL changes every time you start cloudflared. You must update it in YesBill Settings each session.

Step 6 — Configure YesBill​

  1. Open YesBill → Settings → AI
  2. Select the Ollama provider tile
  3. Paste your https://xxxx.trycloudflare.com URL in the Ollama Base URL field
  4. Click Fetch Models — your installed models will appear
  5. Select a model from the list
  6. Click Save Changes

The selected model is now used in Ask AI and the YesBill Agent.

Every Session Checklist​

Each time you restart your computer or want to use Ollama with YesBill:

  • Quit the Ollama tray app (if it auto-started)
  • Run: set OLLAMA_HOST=0.0.0.0:11434 && set OLLAMA_ORIGINS=* && "C:\Users\%USERNAME%\AppData\Local\Programs\Ollama\ollama.exe" serve
  • Run cloudflared in a new window: cloudflared tunnel --url http://localhost:11434
  • Copy the new tunnel URL → paste in YesBill Settings → Fetch Models → Save

Troubleshooting​

"Cannot connect to Ollama"​

  • Make sure Ollama is running (Step 3)
  • Make sure cloudflared is running (Step 5)
  • Make sure the tunnel URL in Settings matches the one printed by cloudflared
  • The tunnel URL changes each session — update it in Settings

"Ollama returned 403"​

Ollama is blocking the request. Run it with OLLAMA_HOST=0.0.0.0 and OLLAMA_ORIGINS=* as shown in Step 3.

"bind: Only one usage of each socket address"​

Another Ollama process is using port 11434. Kill it:

taskkill /F /IM ollama.exe /T
taskkill /F /IM "ollama app.exe" /T

Then retry Step 3.

"Unable to reach the origin service"​

Cloudflare can't reach Ollama. Make sure Ollama started with OLLAMA_HOST=0.0.0.0:11434 (not 127.0.0.1). Restart both services.

No models appear after Fetch Models​

You may not have any models installed. Run:

ollama pull llama3.2

Then click Fetch Models again.

Make it easier

Create a batch file start-ollama.bat on your Desktop with:

@echo off
taskkill /F /IM ollama.exe /T 2>nul
taskkill /F /IM "ollama app.exe" /T 2>nul
timeout /t 1 >nul
start "Ollama" cmd /k "set OLLAMA_HOST=0.0.0.0:11434 && set OLLAMA_ORIGINS=* && C:\Users\%USERNAME%\AppData\Local\Programs\Ollama\ollama.exe serve"
timeout /t 3 >nul
start "Cloudflared" cmd /k "C:\Program Files (x86)\cloudflared\cloudflared.exe tunnel --url http://localhost:11434"

Double-click it to start both services at once. Copy the tunnel URL from the cloudflared window into YesBill Settings.