| Connected Account | The Connected Account with your Google Gemini API key | Connected Account | Yes | |
| Prompt | The prompt to generate completions for. Input will be truncated if it exceeds the limit set by the API | Text (Long) | Yes | |
| Model | The Gemini model to use. Options: Gemini 3.1 Pro Preview, Gemini 3 Flash Preview, Gemini 3.8 Flash | Predefined Choice List | Yes | |
| System message | The message which sets the context and expectations for a conversation with the AI | Text (Long) | No | |
| Maximum output tokens | The maximum number of reasoning and output tokens that can be used in this Wrk Action. If blank, the model default maximum is used. If the Generated text output is blank, try increasing this value, decreasing the thinking level, or turning off Include thoughts? | Number | No | |
| Temperature | Sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. Shown for Gemini 3 Flash Preview and Gemini 3.1 Pro Preview; not supported on Gemini 3.8 Flash | Number with decimals | No | |
| Thinking level | Constrains effort on thinking for thinking models. Reducing thinking effort can result in faster responses and fewer tokens used on thinking. Leave blank to use the model default. Options: Low, Medium, High | Predefined Choice List | No | |
| Files | Files to be provided to the AI | List of files | No | |
| Include thoughts? | If set to true, the response will include the thoughts of the AI if available | True/False | No | |
| Return response as JSON? | If set to true, the response of the AI will return valid JSON | True/False | No | |
| Structured output schema | JSON Schema to enforce structured output. When provided, Gemini will guarantee the response conforms to this schema. Shown when Return response as JSON? is true | Text (Long) | No | |