Chat Completions#
Performs a single stateless Chat Completions request and returns the model's response without creating a chat, message, or launch.
| InTable | [Data table | array of objects] Prepared OpenAI message history; compatible input and If specified, takes precedence over simple text inputs. |
| ContentSystem | [Text] Optional system prompt; the client creates a message with role=system. |
| ContentUser | [Text] User text; required if InTable does not contain prepared messages. The client creates a message with role=user. |
| Model | Model provider selection. |
| CustomModelName | [Text] Optional name of the Sherpa AI Server model; if empty, the model selects the server. |
| Stop | [Text | array of texts] One or more stop sequences for generation. |
| Temperature | [Number] Generation temperature; default value is 0.7. |
| TopP | [Number] Threshold for nucleus sampling; empty value is not sent. |
| MaxLength | [Number] Maximum number of tokens in the response; compatible alias of the API property MaxTokens, default value is 1000. |
| AutoLengthLimit | Automatically limit the response length to the available model context. |
| Tools | [Array of objects] Descriptions of available model tools in OpenAI format. |
| ToolChoice | [Text | object] Tool selection rule: automatic mode, prohibition, or specific tool. |
| Timeout | [Object] Required format for structured response. |
| StructuredOutput | [Number] Timeout for waiting for the model's response in seconds; default value is 300. |
| ResponseRole | [Text] Role of the message from the first response variant. |
| ResponseContent | [Text | object] Content of the message from the first response variant. |
| TotalTokens | [Number] Total number of tokens used from usage.total_tokens. |
| FinishReason | [Text] Reason for the completion of the first variant, such as stop, length, or tool_calls. |
| ToolCalls | [Array of objects] Tool calls from the model's message. |
| Response | Complete typed response Chat Completions. |
| Usage | Usage data for input, output, and all tokens. |
| Error Level | Select error handling level. Possible values:
If "Default" is selected, the value from the "Start" block of this diagram will be used. |
| Message Level | Select the level of messages that will be output by the blocks during operation. Possible values:
If "Default" is selected, the value from the "Start" block of this diagram will be used. |
| Error Text | [Text] Returns detailed information about the error in case of incorrect execution of the block's work. |