Use ScriptEngine in Open WebUI.

Connection scope

Start with the provider path Open WebUI documents

Open WebUI documents OpenAI-compatible APIs as a provider connection: enter a URL and API key, then let the interface detect the models. ScriptEngine exposes the same base URL and bearer-authenticated model discovery pattern, so the first setup is straightforward.

This guide intentionally validates chat only. ScriptEngine publicly verifies /v1/models and /v1/chat/completions; embeddings, image generation, audio, Open Responses, and tool calls must not be assumed from the fact that Open WebUI has controls for them. If you need RAG embeddings, configure a separately verified embedding provider rather than pointing Open WebUI at an unverified route.

Use the official Open WebUI provider guide alongside this page. Menu labels can move between releases, but the connection values remain the same.

Prepare

Discover a model before opening the admin panel

Create a ScriptEngine account, generate a private API key, and query the model list with that exact key. This prevents a common Open WebUI failure: a model name copied from another provider or an outdated tutorial.

export SCRIPTENGINE_API_KEY="YOUR_PRIVATE_KEY"
curl https://scriptengine.org/v1/models \
  -H "Authorization: Bearer $SCRIPTENGINE_API_KEY"

Save one returned model ID as MODEL_ID_FROM_DISCOVERY. The public ScriptEngine model catalog helps you compare current credit rates, but availability for your key is confirmed by the authenticated response.

Open WebUI settings

Add a new OpenAI-compatible connection

1. Sign in to your Open WebUI administrator account.
2. Open Admin Settings → Connections.
3. Choose the OpenAI-compatible provider or the OpenAI connection form, depending on your release.
4. Set the connection URL to https://scriptengine.org/v1.
5. Paste the ScriptEngine API key into the API key field and save.
6. Refresh the model list and select the returned ScriptEngine model.

Do not enter the homepage URL, the version root without /v1, or a full endpoint such as /v1/chat/completions. Open WebUI appends the request path itself.

Verify

Confirm a normal chat response

Send a short message in a new Open WebUI chat. Ask the model to return a fixed phrase such as “ScriptEngine connected.” If the direct cURL request works but Open WebUI fails, compare the URL, key, and selected model field character-for-character.

curl https://scriptengine.org/v1/chat/completions \
  -H "Authorization: Bearer $SCRIPTENGINE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"MODEL_ID_FROM_DISCOVERY","messages":[{"role":"user","content":"Reply with exactly: ScriptEngine connected"}]}'

Known limits

Do not confuse a chat connection with every Open WebUI feature

Open WebUI supports OpenAI-compatible APIs, Open Responses, Ollama, and local servers. ScriptEngine is not claiming all of those protocols. The current public ScriptEngine setup is a bearer-authenticated chat-completions route with model discovery.

• Embeddings: configure a separately verified embedding endpoint for knowledge-base or RAG features.
• Images and audio: do not select those features unless the specific model and endpoint have been tested.
• Tools and functions: Open WebUI controls do not prove that the ScriptEngine route accepts tool schemas.
• Usage and cost: review the ScriptEngine workspace and current model catalog; the 86% figure is a dated reference maximum, not a universal discount.
• Reliability: ScriptEngine currently publishes no uptime or response-time SLA.

Troubleshooting

Resolve the errors that look like provider failures

No models appear

Check that the connection URL ends in /v1, the key is valid, and the Open WebUI server can reach the public endpoint. Run the cURL model-list request from the same machine.

401 unauthorized

Generate or copy the ScriptEngine API key again. Do not use your ScriptEngine account password or a key from another provider.

Model not found

Run /v1/models again and use an ID from that response. The public catalog and another account may show a different current route.

Chat works but RAG does not

That usually indicates a missing or incompatible embeddings connection. Keep chat on ScriptEngine and configure an embedding provider whose endpoint has been tested for your Open WebUI version.

Sources

Keep the setup tied to primary documentation

This guide was checked August 4, 2026 against the official Open WebUI connection documentation and ScriptEngine public API documentation. Open WebUI releases may move menu labels; the protocol and connection fields are the important parts.

• Open WebUI: Connect a provider
• Open WebUI API reference
• ScriptEngine API integration overview
• ScriptEngine model catalog and rates