Prerequisites
1
RuAPI account and API key
Console → API Keys → create an
sk-... key. The balance tops up in USDT — see Top-up.2
A running Open WebUI
Install per the Open WebUI docs (usually Docker).
Option 1: through the UI
1
Open connection settings
Admin Panel → Settings → Connections (or Settings → Connections for the current user).
2
Add an OpenAI connection
Enable the OpenAI API and enter:
Save. Open WebUI pulls the model list via
/v1/models.Option 2: through environment variables
When starting the container, set:Verify
Start a new chat, pick a model (e.g.claude-opus-5-5) and send a message. A reply plus an entry in RuAPI Console → Usage Logs means it works.
Knowledge base (RAG)
To have Open WebUI compute document embeddings through RuAPI: Admin Panel → Settings → Documents → Embedding Model Engine — choose OpenAI, enterhttps://www.ruapi.ai/v1, your RuAPI key and the model text-embedding-3-small (or text-embedding-3-large), and save. The same via environment variables: RAG_EMBEDDING_ENGINE=openai, RAG_OPENAI_API_BASE_URL=https://www.ruapi.ai/v1, RAG_OPENAI_API_KEY=sk-YOUR_KEY, RAG_EMBEDDING_MODEL=text-embedding-3-small (the URL and key can be omitted — then OPENAI_API_BASE_URL and OPENAI_API_KEY are used). The variables apply only on first start; after that, change the settings in the Admin Panel. After changing the embedding model, re-index documents you uploaded earlier. Embedding models unlock after a top-up.
Troubleshooting
The model list is empty
The model list is empty
- Check the Base URL is
https://www.ruapi.ai/v1with/v1: models are fetched via/v1/models. - The key must be a RuAPI key (
sk-...).
401 / auth error
401 / auth error
The key is wrong or has spaces. Check its state in RuAPI Console → API Keys.
I want to keep both Ollama and RuAPI
I want to keep both Ollama and RuAPI
You can: Open WebUI supports several connections at once. Local Ollama models and cloud RuAPI models show up in the same picker.