---
title: "Ask Persona Question Stream V2"
method: POST
path: "/v2/personas/questions/stream"
tags: ["personas-v2"]
---

# Ask Persona Question Stream V2

`POST /v2/personas/questions/stream`

## Request

- Base URL: `https://api.evidenso.com` (the document declares no server; this is the origin it was published from)
- URL: `https://api.evidenso.com/v2/personas/questions/stream`
- Auth: HTTP bearer

V2: Ask a question and receive streaming updates with LangGraph orchestration.

Differences from v1:
- LLM-based orchestrator generates execution plan dynamically
- Skips unnecessary nodes based on question complexity
- Emits orchestrator plan as first event
- Same output format (backward compatible)

Returns:
    StreamingResponse with SSE events:
    - plan: Orchestrator's execution plan
    - status: Progress updates
    - error: Error events (critical/recoverable)
    - result: Final response

## Request body

- QuestionRequestV2: V2 request model for questions endpoint. Same fields as v1 PersonaAskRequest for backward compatibility.
  - `personaId` string, required
  - `question` string, required
  - `conversationId` string, nullable
  - `topK` integer
  - `images` string[], nullable
  - `category` string, nullable
  - `model` string, nullable: LLM model to use (e.g., 'gpt-5.6-sol', 'claude-opus-5'). Defaults to the heavy tier (model_config.TIERS) if not provided
  - `mentionedProducts` SrcV2ModelsRequestsMentionedProduct[], nullable
    - `id` string, required
    - `name` string, required
  - `mentionedFeatures` SrcV2ModelsRequestsMentionedFeature[], nullable
    - `id` string, required
    - `name` string, required
    - `productId` string, required

## Response `200`

Successful Response

- unknown

## Other responses

- `422`: Validation Error

---

[API](https://skmtc.dev/evidenso/apis/persona-rag-api.md) - [All operations](https://skmtc.dev/evidenso/apis/persona-rag-api/llms.txt) - [OpenAPI document](https://skmtc.dev/evidenso/apis/persona-rag-api/revisions/86408c178d4f?raw)