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

# Error messages

> Every message the chat can show when a turn fails, what it means, and where the fix is.

The chat shows one short line when a turn fails. This page maps each line to its cause and to
whoever can fix it: you, your workspace admin, your organisation's IT, or Pretty.

## Your organisation's side

These come from the provider or proxy your workspace connected in
[Model providers](/admin/model-providers). Pretty relays them; it cannot fix them.

| Message                                                            | Meaning                                                                                 | Details                                                        |
| ------------------------------------------------------------------ | --------------------------------------------------------------------------------------- | -------------------------------------------------------------- |
| **Your company's AI service isn't available right now.**           | The connected AI service (LiteLLM proxy or the model provider behind it) did not answer | [AI service isn't available](/errors/ai-provider-unreachable)  |
| **The selected model isn't available on your LiteLLM proxy.**      | The chat's model is not on the proxy's model list                                       | [Troubleshooting](/help/troubleshooting#while-the-agent-works) |
| **The AI provider rejected your API key.**                         | The connected key is invalid, expired or lacks access to the model                      | [Troubleshooting](/help/troubleshooting#while-the-agent-works) |
| **The AI provider's quota has been exceeded.**                     | The account behind the key is out of credit                                             | [Troubleshooting](/help/troubleshooting#while-the-agent-works) |
| **The AI provider is rate limiting requests.**                     | Too many requests too quickly; wait and retry                                           | [Troubleshooting](/help/troubleshooting#while-the-agent-works) |
| **Your organization requires an API key, but none is configured.** | Workspace is set to bring-your-own-key and no key has been added                        | [Model providers](/admin/model-providers)                      |

## Your workspace's side

| Message                                     | Meaning                                                        | Details                                                        |
| ------------------------------------------- | -------------------------------------------------------------- | -------------------------------------------------------------- |
| **You've reached your monthly AI limit.**   | Your personal monthly allowance is spent                       | [Usage and limits](/admin/usage-limits)                        |
| **This chat is busy with another request.** | A turn is already running in this chat; your message is queued | [Troubleshooting](/help/troubleshooting#while-the-agent-works) |

## This chat

| Message                                                                  | Meaning                                                      | Details                                                        |
| ------------------------------------------------------------------------ | ------------------------------------------------------------ | -------------------------------------------------------------- |
| **This chat has grown too large to continue.**                           | Nothing left to compact; start a new chat                    | [Context and compaction](/chat/context)                        |
| **Your message or this chat is too large to continue.**                  | One message or attachment exceeds what a turn can hold       | [Context and compaction](/chat/context)                        |
| **Pretty couldn't safely shorten this chat.**                            | Compaction failed; your messages are saved, start a new chat | [Context and compaction](/chat/context)                        |
| **The AI got stuck repeating the same action, so the turn was stopped.** | Loop guard fired; rephrase more concretely                   | [Troubleshooting](/help/troubleshooting#while-the-agent-works) |

## Pretty's side

| Message                                                  | Meaning                                        | What to do                                                     |
| -------------------------------------------------------- | ---------------------------------------------- | -------------------------------------------------------------- |
| **The connection was interrupted.**                      | The stream between the pane and Pretty dropped | Retry. Pretty retries once on its own first.                   |
| **This app version is out of date for the chat server.** | The pane is older than the server expects      | Click **Reload**.                                              |
| **Something went wrong. Please try again.**              | An error Pretty could not classify             | Retry once, then [tell us](/help/support) with the message ID. |