---
url: /relyai-acellemail/guides/connect-an-engine.md
description: >-
  Add an OpenAI, Anthropic or Ollama engine to RelyAI — create it, test the
  connection, set a default, and decide between a system key and letting
  customers bring their own.
---

# Connect an engine

Add an AI **engine** (provider + model + key), test it, and set a default. You can add several engines and switch the default at any time.

**Prerequisites:** an **Active** license and an API key from your chosen provider.

::: info Engine architecture
An engine pairs a specific provider with a model (e.g. OpenAI GPT-4o mini):

* **Default engine:** Receives and processes all incoming AI requests.
* **Standby engines:** Additional enabled engines remain available for immediate promotion.
  :::

![The Engines & Keys settings tab](/relyai-acellemail/img/relyai-connect-an-engine-list.png)
*RelyAI → Settings → Engines & Keys*

## 1. Add an engine

**RelyAI → Settings → Engines & Keys → Add Engine.** Fill in:

| Field | Notes |
| --- | --- |
| **Name** | Label you'll recognise, e.g. "OpenAI GPT-4o" |
| **Provider** | OpenAI, Anthropic or Ollama |
| **Model** | Model ID for that provider (see below) |
| **Max Tokens** | Response length ceiling |
| **API key** | Placeholder `sk-…` / `sk-ant-…` / *(leave blank for Ollama)*; leaving blank on edit keeps the existing key |

## 2. Get a key for your provider

| Provider | Where to get the key | Models |
| --- | --- | --- |
| **OpenAI** | `platform.openai.com` → **API keys → Create new secret key** → copy `sk-…` | `gpt-4o-mini` (fast/low-cost default), `gpt-4o`, `gpt-5`, `gpt-5-mini`, `gpt-3.5-turbo` |
| **Anthropic** | `console.anthropic.com` → **Settings → API keys** → copy `sk-ant-…` | Claude Opus, Sonnet, Haiku families; `claude-haiku-4-5` a good low-cost default |
| **Ollama** | Start locally with `ollama serve`, set **Base URL** (e.g. `http://127.0.0.1:11434`), leave the key blank; for Ollama Cloud paste the dashboard key | `llama3.1:8b` (good default), `llama3.1:70b`, `mistral:7b`, `qwen2.5:7b`, `qwen3:8b` |

Ollama runs open-weight models with **zero per-token cost when self-hosted**.

## 3. Test the connection

Click **Test Connection**:

* **Success:** Displays *"Testing…"* then *"Connection successful"*.
* **Troubleshooting:** Verify API key syntax, model IDs, and Base URL accessibility (for Ollama).

## 4. Set a default

Click **Set as Default** on your primary engine:

* **Routing:** All AI operations route immediately to the designated default engine.
* **Standby engines:** Other engines remain saved and available for quick switching.

## System key vs customer keys

Configure API billing ownership:

* **System key:** Set on the engine and used by default for all user requests.
* **Customer keys:** Enable **Allow customers to use their own API key** so users enter personal keys under **Account → AI** and consume their own provider quotas. See [Quotas, limits and credits](/relyai-acellemail/guides/quotas-limits-and-credits).

## Related pages

* [Configuration](/relyai-acellemail/configuration)
* [Settings reference](/relyai-acellemail/settings)
* [Quotas, limits and credits](/relyai-acellemail/guides/quotas-limits-and-credits)