Sosyabot Docs

Appearance

Inline Assistance ​

Inline assistance is the synchronous text-generation endpoint — the same one prompt runs hit. It handles single-shot operations like rewriting a caption, translating a draft, shortening to a target length, or expanding a fragment into a full paragraph.

Endpoint ​

POST /api/v1/ai-generate/text

Request body:

typescript
{
  modelId?: string,
  prompt: string,
  options?: {
    language?: string,
    tone?: "casual" | "formal" | "engaging" | "professional",
    creativity?: number,
    hashtags?: boolean | "enable" | "disable",
    words?: number,                // target word count
    systemPromptExtra?: string,    // appended to the synthesized system prompt
  }
}

The service synthesizes a system prompt from options, calls the Qevron gateway, and returns the generated string. If modelId is omitted, the first active model in Models is used.

Token budget ​

If options.words is supplied, max tokens is set to words × 1.7 (the rough English token-to-word ratio). Otherwise the default is 512.

Cost ​

1 credit per call. Debited before the model call.

Suggestions

Prompts

Credits