Custom LLM Settings — Bring Your Own AI Model

Custom LLM Settings — Bring Your Own AI Model

By default, Botaura uses its own LLM (AI model) to power your chatbot. But if you have a specific model preference or an API key from another provider, you can configure your own.

Plan requirement: Custom LLM configuration requires the Pro plan or above.

---

How LLM Resolution Works

Botaura uses a 3-tier fallback system for AI responses:

1. Your custom LLM — if you've configured one, it's tried first
2. Botaura's platform LLM — used as fallback if your custom model fails or isn't configured
3. Hardcoded fallback — a last-resort model that's always available

This means your bot never goes silent — even if your custom API key expires or the provider has downtime, Botaura's own models take over seamlessly.

---

Setting Up a Custom LLM

1. Go to Dashboard → Settings → LLM
2. Choose your provider (e.g., OpenRouter)
3. Enter your API key
4. Select a model (or enter a custom model ID)
5. Click Test Connection to verify it works
6. Save

Security

Your API key is AES-256 encrypted before being stored in our database. We never store raw API keys. The key is only decrypted at runtime when generating a response.

Platform Fallback

Enable "Use Botaura's LLM as fallback" (recommended). If your custom model fails for any reason, the response will be generated by Botaura's platform model instead of showing an error to your customer.

---

Testing

After configuring, click Test Connection. This sends a sample request to your model and measures latency. You'll see:

• Success/failure status
• Response latency (in milliseconds)
• Last test date

---

Removing Custom Config

To switch back to Botaura's default model, click Remove Custom Configuration. Your encrypted API key is deleted immediately.