> Documentation index: https://docs.scrapier.io/llms.txt
> Auth: header `X-API-Key: <key>` · Base: https://api.scrapier.io/api/v2 · Synchronous; paginate with `cursor` ← `data.next_cursor`.

# Instagram Transcript

Transcribe the speech in an Instagram reel or video post from its URL, with detected language and the post's like and comment counts.

`GET|POST https://api.scrapier.io/api/v2/instagram-transcript/` · 2 credits per successful call · Single lookup · Long-running

## Overview

Send a reel or post URL in `urls`. You get back a `transcript` object with the spoken text, detected language, shortcode, author, like and comment counts and a status.

Tune transcription with `model_size` (tiny to large-v3), `beam_size` (1–10), `vad_filter` (drops non-speech audio) and `language` (force a language instead of auto-detect). Set `translate_to_english` to get English output.

## Use cases

- Turn reels into searchable text.
- Summarize or caption video content.
- Research what creators say about a product.

## Request

```bash
curl -G "https://api.scrapier.io/api/v2/instagram-transcript/" \
  -H "X-API-Key: YOUR_API_KEY" \
  --data-urlencode "urls=https://www.instagram.com/reel/DXtvzz7B3dC/" \
  --max-time 210
```

```python
import requests

resp = requests.get(
    "https://api.scrapier.io/api/v2/instagram-transcript/",
    headers={"X-API-Key": "YOUR_API_KEY"},
    params={"urls": "https://www.instagram.com/reel/DXtvzz7B3dC/"},
    timeout=210,
)
body = resp.json()
if not resp.ok:
    raise RuntimeError(body.get("error") or body.get("detail"))

print(body["data"]["transcript"])
```

POST with a JSON body works too (same parameters; the body is merged over the query string).

## Parameters

| Name | Type | Required | Default | Description |
|---|---|---|---|---|
| `X-API-Key` (header) | string | yes | — | Your API key. |
| `urls` | string | yes | e.g. `https://www.instagram.com/reel/DXtvzz7B3dC/` | Reel/post URL. Lookup endpoint — next_cursor null. |
| `translate_to_english` | boolean | no | `false` | Translate transcript to English |
| `model_size` | string | no | `base` | faster-whisper model size. One of: `tiny`, `base`, `small`, `medium`, `large-v3`. |
| `beam_size` | integer | no | `1` | Beam search width (1-10) |
| `language` | string | no | — | ISO language code to force (default: auto-detect) |
| `max_duration_seconds` | integer | no | `0` | Optional cap: videos longer than this many seconds are rejected up front (422). 0 = no cap (full transcript, or a typed 504 if it cannot finish in time) |
| `vad_filter` | boolean | no | `true` | Drop non-speech audio |
| `use_proxy` | boolean | no | `true` | Route the request through Scrapier's residential proxy pool. Default: `true`. Leave as is unless support tells you otherwise. |
| `raw_data` | boolean | no | `false` | When true, the response carries the full raw upstream payload(s) the scraper received (GraphQL / JSON API bodies, HTML, embedded JSON-LD -- exactly as the platform returned them) under `raw_responses`, in place of the formatted items. When false (default) you get the formatted output. Pagination (`cursor` / `next_cursor`) works the same in both modes. |

## Response

HTTP 200. Results are in `data.transcript`. Sample from the live catalog, arrays trimmed to 3 items.

```json
{
  "success": true,
  "data": {
    "urls": "https://www.instagram.com/reel/DXtvzz7B3dC/",
    "transcript": {
      "input_url": "https://www.instagram.com/reel/DXtvzz7B3dC/",
      "transcript": "Welcome to Survivor 50 Mr. Beast. Come on, baby...",
      "language": "en",
      "code": "DXtvzz7B3dC",
      "user_name": "survivorcbs",
      "like_count": 64469,
      "comment_count": 6738,
      "status": "success"
    },
    "next_cursor": null
  }
}
```

## Response fields

| Field | Type | Example | Description |
|---|---|---|---|
| `urls` | string | `https://www.instagram.com/reel/DXtvzz7B3dC/` |  |
| `transcript` | object | `{…}` | The results. |
| `next_cursor` | null | `null` | Pass back as `cursor` for the next page. `null` on the last page. |
| `transcript.input_url` | string | `https://www.instagram.com/reel/DXtvzz7B3dC/` |  |
| `transcript.transcript` | string | `Welcome to Survivor 50 Mr. Beast. Come on, ba…` |  |
| `transcript.language` | string | `en` |  |
| `transcript.code` | string | `DXtvzz7B3dC` |  |
| `transcript.user_name` | string | `survivorcbs` |  |
| `transcript.like_count` | integer | `64469` |  |
| `transcript.comment_count` | integer | `6738` |  |
| `transcript.status` | string | `success` |  |

## Pagination

Single lookup: everything comes back in one call and `next_cursor` is always `null`.

## Errors

Failed calls are free. Branch on HTTP status; the message is in `error` (or `detail`). See https://docs.scrapier.io/errors.md

- **400** Bad input. The exact message varies by endpoint. `{"success":false,"error":"…"}`
- **402** Balance below the endpoint's price. `{"error":"Insufficient credits. Buy more to scrape more.","balance":1,"required":2}`
- **403** Invalid or unknown API key. `{"detail":"Authentication service unavailable or invalid API key."}`
- **429** Throttled — wait for Retry-After. `{"detail":"Request was throttled. Expected available in N seconds."}`
- **404** unknown slug · **503** busy (retry after `Retry-After`) · **5xx** upstream failure or timeout (retry with backoff).

## Notes

- `model_size` accepts tiny, base, small, medium or large-v3. The default is base.
- Check `status` in the result: it reports whether transcription succeeded.
- This endpoint can take several minutes. Set your client timeout to at least 210 seconds.

## Related

- [Instagram Reel](https://docs.scrapier.io/api-reference/instagram/instagram-reel-scraper.md): Fetch one Instagram reel by URL or shortcode, with caption, owner, view, like and comment counts and its latest comments.
- [Instagram Post](https://docs.scrapier.io/api-reference/instagram/instagram-post-scraper.md): Fetch one Instagram post by URL or shortcode, with caption, owner, like and comment counts and its latest comments.
- [Instagram Profile Reels](https://docs.scrapier.io/api-reference/instagram/instagram-profile-reels.md): List a public account's reels, a page at a time, with caption, like and comment counts and posting time, from the username.
