> ## Documentation Index
> Fetch the complete documentation index at: https://docs.submagic.co/llms.txt
> Use this file to discover all available pages before exploring further.

# Edit the Transcript Before Rendering

> Create a project without auto-rendering, correct the captions, then export the final video

# Edit the Transcript Before Rendering

By default, a project created through the API is **rendered automatically** as
soon as transcription finishes. If you want to review or fix the captions first
(correct a misheard word, retime a caption, add or remove words), create the
project with `autoRender: false`. The project then waits for you to edit the
transcript and trigger the export yourself.

<Steps>
  <Step title="Create the project with autoRender: false">
    Pass `autoRender: false` to [Create Project](/api-reference/create-project).
    The video is downloaded and transcribed, but **no render is started**.

    ```bash cURL theme={null}
    curl -X POST "https://api.submagic.co/v1/projects" \
      -H "x-api-key: sk-your-api-key-here" \
      -H "Content-Type: application/json" \
      -d '{
        "title": "My Awesome Video",
        "language": "en",
        "videoUrl": "https://example.com/videos/sample.mp4",
        "templateName": "Hormozi 2",
        "autoRender": false,
        "webhookUrl": "https://yoursite.com/webhook/submagic"
      }'
    ```
  </Step>

  <Step title="Wait until transcription is ready">
    You have two options to learn when the transcript is ready:

    * **Webhook (recommended):** if you passed a `webhookUrl`, you'll receive a
      `completed` notification once transcription finishes. For an
      `autoRender: false` project this fires **after transcription** (the video
      is not rendered yet, so `downloadUrl` will be empty until you export).
    * **Polling:** call [Get Project](/api-reference/get-project) until
      `transcriptionStatus` is `"COMPLETED"`. At that point the `words` array is
      available.

    ```bash cURL theme={null}
    curl "https://api.submagic.co/v1/projects/{id}" \
      -H "x-api-key: sk-your-api-key-here"
    ```

    ```json Response (transcription done) theme={null}
    {
      "id": "550e8400-e29b-41d4-a716-446655440000",
      "status": "completed",
      "transcriptionStatus": "COMPLETED",
      "words": [
        { "id": "w1EydrN5", "text": "Do",  "type": "word", "startTime": 0.08, "endTime": 0.16 },
        { "id": "8VUdYfGN", "text": "you", "type": "word", "startTime": 0.16, "endTime": 0.24 }
      ]
    }
    ```

    <Note>
      For an `autoRender: false` project, `status: "completed"` means
      "transcribed and ready to edit" — the video has **not** been rendered yet.
      A `downloadUrl` only appears after you export in step 4.
    </Note>
  </Step>

  <Step title="Get the words array and edit it">
    Take the `words` array from the Get Project response and change whatever you
    need. You can:

    * **Fix text** — change a word's `text`.
    * **Retime** — adjust `startTime` / `endTime`.
    * **Add a word** — insert a new entry and **omit its `id`** (a fresh id is
      assigned for you).
    * **Remove a word** — leave it out of the array.

    Send the **full** array back — it replaces the current transcript.
  </Step>

  <Step title="Submit the edited words">
    Send the edited array to [Update Project](/api-reference/update-project) in
    the `words` field. The caption layout is rebuilt automatically so everything
    stays in sync at render time.

    ```bash cURL theme={null}
    curl -X PUT "https://api.submagic.co/v1/projects/{id}" \
      -H "x-api-key: sk-your-api-key-here" \
      -H "Content-Type: application/json" \
      -d '{
        "words": [
          { "id": "w1EydrN5", "text": "Did", "type": "word", "startTime": 0.08, "endTime": 0.16 },
          { "id": "8VUdYfGN", "text": "you", "type": "word", "startTime": 0.16, "endTime": 0.24 }
        ]
      }'
    ```

    ```json 200 OK theme={null}
    {
      "message": "Project updated successfully",
      "id": "550e8400-e29b-41d4-a716-446655440000",
      "status": "completed"
    }
    ```

    <Note>
      Editing words is only allowed **after** transcription completes. Calling it
      too early returns a `VALIDATION_ERROR`.

      You can also combine `words` with other fields (e.g. `music`, `items`) in
      the same request — the caption edit is applied first, then the rest.
    </Note>
  </Step>

  <Step title="Export the final video">
    When the captions look right, trigger the render with
    [Export Project](/api-reference/export-project). You'll be notified at your
    `webhookUrl` (or can poll Get Project) when the rendered video is ready and
    `downloadUrl` is populated.

    ```bash cURL theme={null}
    curl -X POST "https://api.submagic.co/v1/projects/{id}/export" \
      -H "x-api-key: sk-your-api-key-here" \
      -H "Content-Type: application/json" \
      -d '{}'
    ```
  </Step>
</Steps>

## Notes

* You can edit the transcript as many times as you like before exporting —
  repeat steps 3–4.
* Default behavior is unchanged: omit `autoRender` (or set it to `true`) and the
  project renders automatically once transcription completes, exactly as before.
* This flow does not apply to Magic Clips, which follow their own pipeline.