You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 3aa250a
Browse filesBrowse the repository at this point in the historyBrowse files
description: Configure AMD Lemonade as a custom endpoint in LibreChat.
5
+
---
6
+
7
+
AMD Lemonade runs models locally through an OpenAI-compatible API. Follow the [Lemonade documentation](https://lemonade-server.ai/docs/) to install and start the server, then download a chat model.
8
+
9
+
## Configuration
10
+
11
+
Lemonade requires no API key; the placeholder below satisfies the OpenAI client. Add the endpoint under `endpoints.custom` in your `librechat.yaml`:
12
+
13
+
```yaml filename="librechat.yaml"
14
+
- name: 'lemonade'
15
+
apiKey: 'lemonade'
16
+
baseURL: 'http://localhost:13305/v1'
17
+
models:
18
+
default: ['Qwen3-0.6B-GGUF']
19
+
fetch: true
20
+
titleConvo: true
21
+
titleModel: 'current_model'
22
+
modelDisplayLabel: 'AMD Lemonade'
23
+
iconURL: 'lemonade'
24
+
```
25
+
26
+
## Notes
27
+
28
+
- Replace `Qwen3-0.6B-GGUF` with a locally installed chat model. With `fetch: true`, LibreChat retrieves the model list from Lemonade. `titleModel: 'current_model'` reuses the conversation's model for titles.
29
+
- For Docker, replace `localhost` with `host.docker.internal` or a reachable Lemonade hostname.
30
+
- For user-provided base URLs, add the local server's exact `host:port` (e.g. `host.docker.internal:13305`) to [`endpoints.allowedAddresses`](/docs/configuration/librechat_yaml/object_structure/config#endpoints).
31
+
- Built-in branding recognizes `lemonade`, `AMD Lemonade`, and `Lemonade Server`. Keep `iconURL: 'lemonade'` to preserve the logo when renaming the endpoint.
0 commit comments