Model endpoint
File: config/system/system-model.json. Side: server. A missing file is read after being copied from the file of the same name under the JAR's defaults.
| Field | Type | Required | Default | Description |
|---|---|---|---|---|
model | String | no | "deepseek-chat" | Model name, deepseek-chat by default. |
baseUrl | String | no | "https://api.deepseek.com" | OpenAI-compatible service address, HTTP / HTTPS supported. |
apiKeyEnvironment | String | no | "SYSTEM_DEEPSEEK_API_KEY" | Name of the environment variable holding the API key; not the key itself. |
providerType | String | no | "auto" | auto, openai or deepseek; auto detects it from the service. |
thinkingEnabled | Boolean | no | false | DeepSeek thinking mode option. |
reasoningEffort | String | no | "high" | Requested reasoning effort, high by default; actual support depends on the model. |
apiKeyRequired | Boolean | no | true | Whether the service requires an API key; can be disabled for local services without authentication. |
jsonMode | Boolean | no | true | Whether to use the service's JSON output mode. |
outputTokenParameter | String | no | "max_tokens" | max_tokens or max_completion_tokens. |
Bundled default file
config/system/system-model.json
{
"model": "deepseek-chat",
"baseUrl": "https://api.deepseek.com",
"apiKeyEnvironment": "SYSTEM_DEEPSEEK_API_KEY",
"providerType": "auto",
"thinkingEnabled": false,
"reasoningEffort": "high",
"apiKeyRequired": true,
"jsonMode": true,
"outputTokenParameter": "max_tokens"
}