> ## Documentation Index
> Fetch the complete documentation index at: https://docs.enconvo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Perplexity

> Chat with Perplexity models in Enconvo using your own Perplexity API key.

Perplexity's Sonar models search the web as they answer. Add your key to Enconvo to use those models in chats, Agents, and AI Commands. Perplexity bills the usage to your own account.

## Connect

**1. Get an API key**

Create a key in the [Perplexity API settings](https://www.perplexity.ai/settings/api).

**2. Add it to Enconvo**

Open **Settings → AI Model → Perplexity**. Under **Authentication**, paste the key into **API Key**, then click **Validate**.

**3. Pick a model**

Choose a model in **Model Name** and send a short message. A reply means the connection works.

To keep a second Perplexity key, for example one for work, click **Add New Provider** and create another profile.

## Models and options

| Setting | What it does |
| - | - |
| **Model Name** | The model this provider uses. **Refresh** beside it reloads the list; **Manage Models** (the gear) adds or restores models. |
| **frequencyPenalty** | Lowers the chance of repeated words. The default is 1. |

To use a different model in one chat, pick it from that chat's model picker.

This chat provider is separate from **Settings → Web Search**: choosing Perplexity here doesn't change the search provider your Agents use. When an answer is about current events, open the sources it cites.

## Troubleshooting

| Problem | What to do |
| - | - |
| **Validate** fails | Paste the whole key again and check that it is active in your Perplexity account. |
| A model is missing | Click **Refresh** beside **Model Name** to reload the list. To add a model yourself, use **Manage Models**. |
| Images or tools fail | Pick a model that supports image input or tool calls. |
| Rate or usage limit errors | Check usage and limits in your Perplexity account, or switch to another provider for now. |
| Replies are slow | Try a smaller model, a shorter input, or lower reasoning effort. |

## Related

> [card] [Choose a model](https://docs.enconvo.ai/ai/model-selection-guide.md)
>
> Pick a model by task, and see where defaults are set.

> [card] [Credentials and profiles](https://docs.enconvo.ai/configuration/credentials.md)
>
> Keep several keys, and see what a shared key affects.

> [card] [Auto Route](https://docs.enconvo.ai/providers/auto-route.md)
>
> Let Enconvo pick the model for each message.

> [card] [Provider catalog](https://docs.enconvo.ai/providers/catalog.md)
>
> Every provider, grouped by the feature it powers.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.