Actions18
- Image Actions
- Video Actions
- Audio Actions
- LLM Actions
- Creative Workflow Actions
- Model Actions
- Account Actions
LLM → Generate
AI-generatedSummary
Generate text completions using a specified Sogni large language model (LLM) by providing prompts and optional parameters to customize the output.
Inputs
- Model Search — Optional string to filter available chat models by ID, used to narrow down model selection.
- Model (required) — The Sogni chat or LLM model to use for generation, selected from dynamically loaded options.
- Prompt (required) — The primary user prompt string to send to the selected Sogni LLM model.
- System Prompt — Optional system-level instruction to guide the model's tone, style, or behavior.
- LLM Additional Fields — Collection of optional fields to further customize generation, including Max Tokens (max number of tokens to generate), Messages JSON (full chat messages array in JSON which overrides Prompt and System Prompt), Thinking (boolean to enable reasoning if supported), Use Sogni Hosted Tools (boolean to enable various hosted creative tools), Tools JSON (custom OpenAI-style tools definitions as JSON merged with hosted tools), Tool Choice JSON (controls tool usage mode or specifies a tool choice), and Token Type (which token type to use: Spark or SOGNI).
Output shape
a single record representing the LLM response including the generated text and metadata
The output corresponds to the generated text completion from the chosen Sogni model based on inputs. If Messages JSON is provided, it overrides Prompt and System Prompt. Tool-related fields enable advanced tooling support and interaction. Output structure follows Sogni API chat completion responses.