Forced Alignment

Knox exposes ElevenLabs-compatible forced alignment endpoints at https://api.knoxstudio.ai/v1 for KnoxStudio. Authenticate with your Knox API key (Authorization: Bearer sk-... or xi-api-key). Knox bills the request and relays it upstream. The same routes are also served by https://api.knox.chat/v1.

This is not Speech to Text. Alignment takes audio plus a transcript you already have and returns character- and word-level timestamps.

Typical flow:

1. List models — GET /v1/forced-alignment/models
2. Align — POST /v1/forced-alignment

https://api.knoxstudio.ai/v1
Authorization: Bearer sk-...

Use the official ElevenLabs SDK against Knox by pointing it at this base URL and using your Knox key as xi-api-key.

Models and billing

curl https://api.knoxstudio.ai/v1/forced-alignment/models

model_id	Notes	Price
forced_alignment	Default. Character- and word-level timestamps. Up to 10 hours and 675,000 transcript characters.	$0.27 / audio hour (same published rate as Scribe v2)

Multipart uploads may be up to 3GB. Audio must be at least 100ms. The file and text fields are both required.

Unauthenticated catalog requests are limited to 60/minute and may be cached for 5 minutes. A Bearer token returns only the models that token is allowed to use.

Create an alignment

Send multipart/form-data with the audio file and the transcript text. Optional query: enable_logging.

cURL

curl -X POST https://api.knoxstudio.ai/v1/forced-alignment \
     -H "Authorization: Bearer $KNOXCHAT_API_KEY" \
     -F "file=@speech.wav" \
     -F "text=Hello from Knox. This is forced alignment."

Python

import os
import requests

with open("speech.wav", "rb") as audio:
    response = requests.post(
        "https://api.knoxstudio.ai/v1/forced-alignment",
        headers={"Authorization": f"Bearer {os.environ['KNOXCHAT_API_KEY']}"},
        files={"file": audio},
        data={"text": "Hello from Knox. This is forced alignment."},
    )
response.raise_for_status()
print(response.json()["words"])

A successful 200 includes characters and words arrays with text, start, end, and optional loss. Unknown multipart fields are forwarded upstream.

Errors

422 uses { "detail": [{ "loc": [...], "msg": "...", "type": "value_error" }] }. Other errors use { "detail": "..." }.

Common statuses: 401 missing/invalid key, 402 insufficient credits, 403 model not allowed, 422 missing file / text, over 675,000 characters, or over 10 hours.