---
title: "Fetch the finalized Gladia transcript for a relayed session"
method: GET
path: "/v1/dictation/live/{session_id}"
tags: ["Desktop"]
---

# Fetch the finalized Gladia transcript for a relayed session

`GET /v1/dictation/live/{session_id}`

## Request

- Base URL: `https://api.quillhub.ai`
- URL: `https://api.quillhub.ai/v1/dictation/live/{session_id}`
- Auth: HTTP bearer

Proxies Gladia's result endpoint so the client never needs an API key. Requires the `poll_token` returned by POST /v1/dictation/live — it names the exact key the session was created with (GladiaKeys.id), so this always polls under the right account instead of a freshly-claimed key that likely can't see the session at all. Returns Gladia's payload unchanged.

## Path parameters

- `session_id` string, required

## Query parameters

- `poll_token` string, required

## Response `200`

Gladia payload, passed through.

- unknown

## Other responses

- `401`: Unauthenticated, or poll_token missing/invalid/expired.
- `403`: poll_token was not issued to the authenticated user.
- `429`: Too many requests.
- `502`: Upstream refused.
- `503`: Relay unconfigured, or the key named by poll_token is gone.

---

[API](https://skmtc.dev/quillhub/apis/quillhub-public-api.md) · [All operations](https://skmtc.dev/quillhub/apis/quillhub-public-api/llms.txt) · [OpenAPI document](https://skmtc.dev/quillhub/apis/quillhub-public-api/revisions/5bd42ee9d967?raw)