Open WebUI
Open WebUI is a self-hosted chat interface, usually run in Docker. It treats any OpenAI-compatible URL as a connection.
Setup guide: this config comes from Open WebUI's official documentation and has not been run end to end against the gateway. Corrections to support@hyphen-solution.com.
1. Get a key
app.hyphen-solution.com → sign in → pick a
plan or credit pack → Dashboard → API Keys. Copy the sk-... key.
2a. First install: environment variables
Code
Or in docker-compose.yml:
Code
These variables only apply on first boot
OPENAI_API_BASE_URL and OPENAI_API_KEY are persisted into Open WebUI's
database the first time the container starts. Changing them later in the
compose file does nothing. After the first boot, edit the connection in the
UI instead.
2b. Existing install: the Admin Settings form
Admin Settings → Connections → OpenAI → ➕ Add Connection
- URL:
https://api.hyphen-solution.com/v1 - API Key:
sk-...
Click Save. If the model dropdown stays empty, add the IDs you want to the Model IDs (Filter) allowlist on that connection:
Code
Running Hyphen alongside another provider
Use the plural variables. They are semicolon-separated and paired positionally, so the first URL uses the first key:
Code
3. Verify
Open http://localhost:3000, start a new chat, pick minimax-m3 from the model
selector, and send:
Code
Troubleshooting
- No models in the dropdown. The connection is unreachable, or the filter list is blocking them. Check Admin Settings → Connections and hit the refresh icon on the connection.
- Config changes ignored. You edited environment variables on a container
that has already booted once. Change it in Admin Settings, or delete the
open-webuivolume and start over. 401. Wrong key, or the key is still the placeholder.- Empty replies. Reasoning models need output headroom. See Choosing a model.
429. Monthly budget spent. See Rate limits & caps.- Variable names changed? Check Open WebUI's environment reference.
Related
- Models: the full catalog.
- Choosing a model: which model for which job.