Kerux
Start free
PricingCompareCalculatorContactSign in
Setup · about 2 minutes

Connect your AI to Kerux

Kerux answers tickets with an AI key you bring, so you pay your provider directly, never a markup. There are two ways to do it. Pick one.

Recommended · easiest

A cloud API key

From your preferred provider. Works with Kerux right away, best for almost everyone.

Advanced · private

A local open-source model

Runs on your own machine, nothing leaves your network. A little more setup.

Option 1

Use a cloud API key

Kerux works with any of these 8 providers (plus anything OpenAI-compatible). Pick one, open its console, create a key named “Kerux”, and copy it right away, most providers show it only once. Lost it? Just create another.

GroqFree tier · easiest start
  1. Sign in (or sign up) at console.groq.com.
  2. Go to API Keys → Create API Key.
  3. Copy the key (it starts with gsk_) and paste it into Kerux, see below.

Currently has a free tier with no card required. The fastest way to try Kerux. Paid tiers lift the rate limits.

Anthropic (Claude)
  1. Sign in (or sign up) at platform.claude.com.
  2. Go to API keys → Create key.
  3. Copy the key (it starts with sk-ant-) and paste it into Kerux, see below.

Add prepaid credits to your account, then create a key. Claude models are a strong fit for support replies.

OpenAI (ChatGPT)
  1. Sign in (or sign up) at platform.openai.com.
  2. Go to API Keys → Create new secret key.
  3. Copy the key (it starts with sk-) and paste it into Kerux, see below.

Requires billing set up on your OpenAI account. The mini/nano models cost cents per hundreds of tickets.

Google (Gemini)Free tier
  1. Sign in (or sign up) at aistudio.google.com.
  2. Go to Create API key.
  3. Copy the key (it starts with AIza) and paste it into Kerux, see below.

Currently has a free tier (rate-limited) with a Google account. No card needed to start.

MistralFree tier
  1. Sign in (or sign up) at console.mistral.ai.
  2. Go to API Keys → Create new key.
  3. Copy the key and paste it into Kerux, see below.

Currently offers a free tier after activation, no payment method upfront.

DeepSeek
  1. Sign in (or sign up) at platform.deepseek.com.
  2. Go to API Keys → Create new API key.
  3. Copy the key (it starts with sk-) and paste it into Kerux, see below.

Pay-as-you-go: top up a small balance first. Among the cheapest capable models available.

OpenRouter (400+ models, one key)
  1. Sign in (or sign up) at openrouter.ai.
  2. Go to Keys → Create.
  3. Copy the key (it starts with sk-or-) and paste it into Kerux, see below.

One key routes to 400+ models across every major lab, including some free ones. Handy if you want to experiment before settling.

xAI (Grok)
  1. Sign in (or sign up) at console.x.ai.
  2. Go to API Keys → Create API key.
  3. Copy the key (it starts with xai-) and paste it into Kerux, see below.

Create a team, add billing (xAI sometimes offers starter credits), then create your key.

That's itPaste the key into Kerux (below) and your AI is live. Your AI usage is billed by your provider under their own pricing. At cost, and never marked up or metered by us.
Option 2

Run a local open-source model

Free and fully private. The AI runs on your own computer.

  1. 1
    Install Ollama

    Download it free from ollama.com. Installers for macOS, Windows, and Linux. It runs open-source models on your own machine.

    macOS
    Download the installer from ollama.com, open it, and move the app to Applications. Launch it once.
    Windows
    Download the Windows installer from ollama.com and run it. It sets everything up for you.
    Linux
    Paste Ollama's one-line install script from ollama.com into a terminal.
  2. 2
    Download a model

    Open a terminal and run e.g. “ollama pull llama3.3”. Popular open models include Llama, Qwen, and Mistral. A smaller model is a fast, capable starting point.

  3. 3
    It now serves locally

    Ollama exposes an OpenAI-compatible endpoint at localhost:11434. That's the address Kerux connects to (see the note below about making it reachable).

Good to knowKerux runs in the cloud, so it can't reach a model on your computer's local network directly. To use a local model with hosted Kerux, expose the runner at a secure public address (a tunnel or a small server), then point Kerux at it. If that sounds like a lot, the cloud API key above is the simpler path, and several providers have a free tier to start on.

Paste your key into Kerux

  1. 1. Go to Settings → AI Training.
  2. 2. Choose your AI provider. Any of the 8 above, or “Other” for a local or OpenAI-compatible endpoint.
  3. 3. Paste your key and click Save. It's encrypted, never shown in your browser.
  4. 4. Pick a model, then send a test ticket and watch it reply.
Start free, bring your own key