Chat client setup guide
Plan an OpenAI-compatible client configuration by checking URL handling, key storage, and model identifiers.
Check the client and connection contract
Website under maintenance. This guide describes planned configuration only: api.ubtl.net is a planned domain, no real API test has been performed, and these examples do not establish connectivity. Confirm the supported protocol, request paths, and model identifiers before use.
For Chatbox, NextChat, LobeChat, TypingMind, or Open WebUI, inspect the documentation for your installed version rather than assuming identical menus or compatibility. Check whether its chosen provider supports a custom OpenAI-compatible endpoint. Back up the existing settings or create a separate profile before changing anything.
Map endpoint, key, and model fields
The JSON below is a field-mapping illustration, not a universal configuration or import file. The planned Base URL is https://api.ubtl.net/v1. A client may instead request a host root or append a version and request path automatically; inspect that behavior to avoid a doubled /v1. Do not paste a complete chat request path into every endpoint field.
Replace placeholders only after an authorized key and exact model ID are available. Use an isolated key and check where the application stores it. For a shared deployment, review server-side secret storage and access controls; never place a private key in public frontend settings, exported profiles, or a repository. API access is separate from a ChatGPT web subscription and does not require sharing web passwords or cookies.
{
"provider": "OpenAI-compatible",
"base_url": "https://api.ubtl.net/v1",
"api_key": "YOUR_UBTL_API_KEY",
"model": "YOUR_MODEL_ID"
}Reload and validate minimally
After saving, reopen the application or reload its deployment as required. First inspect the selected provider, assembled URL, and model name without sending a request. Only after service availability and endpoint details are confirmed should you try a short, non-sensitive message. For authentication errors, inspect key loading and permissions; for path errors, inspect URL assembly; for model errors, compare the exact identifier. Redact authorization headers before sharing diagnostics and restore the backup if needed.