How to Transcribe Facebook Videos with Salad Transcription API

Last Updated: April 7, 2025 Facebook URLs are not natively downloadable, which can be a challenge when working with the Salad Transcription API, which requires downloadable audio links. However, you can extract audio from Facebook videos, upload it to Salad Simple Storage Service free of charge, and use the resulting link with the API. This guide will walk you through the process to batch transcribe multiple Facebook videos and save the outputs.

Prerequisites

Before you begin, make sure you have:

Python Installed: Ensure you have Python 3.8 or higher.
Libraries Installed: Use the following command to install the required libraries:
pip install yt-dlp requests
FFmpeg Installed: FFmpeg is needed by yt-dlp for processing audio:
- Linux: sudo apt install ffmpeg
- MacOS: brew install ffmpeg
- Windows: Download FFmpeg and add it to your system PATH.
Salad API Key: Ensure you have a valid Salad API key.

Step 1: Import Libraries

To start, import the required Python libraries:

# For extracting audio from Facebook
import yt_dlp

# For handling HTTP requests and file paths
import requests
from pathlib import Path
import time

# For working with JSON files (batch processing).
import json

Step 2: Define Functions

1. Download Audio from Facebook This function downloads the best-quality audio from a Facebook video:

def download_audio_from_Facebook(video_url, output_format="mp3"):
    """
    Downloads audio from a Facebook video in the specified format.

    Args:
        video_url (str): The URL of the Facebook video.
        output_format (str): Desired audio format (default: "mp3").

    Returns:
        str: The filename of the downloaded audio file.
    """
    try:
        ydl_opts = {
            'format': 'bestaudio/best',
            'postprocessors': [{
                'key': 'FFmpegExtractAudio',
                'preferredcodec': output_format,
                'preferredquality': '128',
            }],
            'outtmpl': '%(id)s.%(ext)s',
        }
        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
            info = ydl.extract_info(video_url, download=True)
            return f"{info['id']}.{output_format}"
    except Exception as e:
        print(f"Error downloading audio: {e}")
        return None

Facebook videos must be in the format https://www.facebook.com/user/videos/name-of-video/38947598374958739457/. Reels, or other video URL formats are not currently supported.

2. Upload Audio to Salad Storage This function uploads the audio file to Salad Storage and generates a presigned URL:

def upload_to_salad_storage(audio_file_path, salad_api_key, organization_name, mime_type="audio/mpeg"):
    """
    Uploads an audio file to Salad Storage and generates a signed access URL.

    Args:
        audio_file_path (str): Path to the audio file.
        salad_api_key (str): Salad API key for authentication.
        organization_name (str): Name of the Salad organization.
        mime_type (str): MIME type of the file (default: "audio/mpeg").

    Returns:
        str: The signed URL for the uploaded file.
    """
    try:
      # Generate the upload URL
      file_name = f"audio/{Path(audio_file_path).name}"
      upload_url = f"https://storage-api.salad.com/organizations/{organization_name}/files/{file_name}"

      # Prepare headers for authentication
      headers = {"Salad-Api-Key": salad_api_key}

      # Prepare data for the upload
      data = {
          "mimeType": mime_type,
          "sign": "true",  # Request a signed URL
          "signatureExp": 3 * 24 * 60 * 60  # 3 days in seconds
      }

      # Prepare the file to be uploaded
      with open(audio_file_path, "rb") as f:
          files = {"file": (Path(audio_file_path).name, f)}
          response = requests.put(upload_url, headers=headers, data=data, files=files)
          response.raise_for_status()

      # Extract and return the signed URL
      signed_url = response.json().get("url")
      if not signed_url:
          raise ValueError("Signed URL not returned by the API.")
      return signed_url

    except requests.exceptions.RequestException as e:
        print(f"Error uploading to Salad storage: {e}")
    except ValueError as ve:
        print(f"API response error: {ve}")
    return None

Step 3: Define Variables

Set the variables required for the process:

# Salad Storage credentials
SALAD_API_KEY = "YOUR_SALAD_API_KEY"  # Replace with your Salad API key
ORGANIZATION_NAME = "YOUR_ORGANIZATION_NAME"  # Replace with your organization name

# List of Facebook URLs
facebook_urls = [
    "https://www.facebook.com/user/videos/name-of-video/38947598374958739457/",
    "https://www.facebook.com/user/videos/name-of-video/39475903749587893745/",
    "https://www.facebook.com/user/videos/name-of-video/29837453749587394857/"
]

Adjust the SALAD_API_KEY and ORGANIZATION_NAME variables to your Salad API key and organization name.
Add all of your Facebook URLs to the Facebook_URLS list.
Facebook videos must be in the format https://www.facebook.com/user/videos/name-of-video/38947598374958739457/. Reels, or other video URL formats are not currently supported.

Step 4: Batch Process Multiple URLs

To process multiple Facebook URLs, use the following script:

# Dictionary to store mappings of Facebook URLs to Salad Storage links
url_mappings = {}

for url in Facebook_URLS:
    try:
        print(f"Processing: {url}")
        audio_path = download_audio_from_Facebook(url)  # Step 1: Download audio
        print(f"Downloaded audio successfully, uploading to Salad Storage...")
        storage_url = upload_to_salad_storage(audio_path, SALAD_API_KEY, ORGANIZATION_NAME)  # Step 2: Upload audio
        url_mappings[url] = storage_url
        Path(audio_path).unlink()  # Clean up local file
        print(f"Processed successfully: {url}")
    except Exception as e:
        print(f"Failed to process {url}: {e}")

# Save mappings to a JSON file
with open("url_mappings.json", "w") as f:
    json.dump(url_mappings, f, indent=4)

print("Batch processing complete. URL mappings saved to url_mappings.json.")

Expected Output

A url_mappings.json file is generated with the mappings:

{
  "https://www.facebook.com/user/videos/name-of-video/29837453749587394857/": "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token=TOKEN",
  "https://www.facebook.com/user/videos/name-of-video/39475903749587893745/": "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token=TOKEN",
  "https://www.facebook.com/user/videos/name-of-video/29837453749587394857/": "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token=TOKEN"
}

Salad Storage supports a max file size of 100MB. If the Facebook video is too long, you may need to split the audio or use a different storage solution.

Step 5: Send Prepared URLs to the Salad Transcription API

Once the audio files are uploaded to Salad Storage and you have the corresponding signed URLs, the final step is to send these URLs to the Salad Transcription API for processing. You can find detailed instructions on how to send requests a to the Salad Transcription API in the official documentation: Salad Transcription API Documentation Below, we provide a function to send a transcription request to the API and retrieve the job_id.

Function to Send a Transcription Request

This function sends a transcription request to the Salad Transcription API for a single URL:

def send_transcription_request(audio_url, organization_name, salad_key, language_code="en", input_params=None):
    """
    Sends a transcription request to the Salad Transcription API with dynamic input parameters.

    Args:
        audio_url (str): The signed Salad Storage URL for the audio file.
        api_endpoint (str): The Salad Transcription API endpoint.
        salad_key (str): The Salad API key for authentication.
        language_code (str): The language of the audio file (default: "en").
        input_params (dict): Additional input parameters to include in the request.

    Returns:
        dict: The response from the API, including the job ID.
    """
    # Set headers
    headers = {
        "Salad-Api-Key": salad_key,
        "Content-Type": "application/json",
        "accept": "application/json",
    }
    api_endpoint = f"https://api.salad.com/api/public/organizations/{organization_name}/inference-endpoints/transcribe/jobs"
    # Prepare the input dictionary, appending the audio URL and language code
    input_data = input_params.copy() if input_params else {}
    input_data["url"] = audio_url
    input_data["language_code"] = language_code

    # Request body
    data = {
        "input": input_data
    }

    # Send the request
    response = requests.post(api_endpoint, headers=headers, json=data)
    response.raise_for_status()  # Raise an error for bad responses

    # Extract and return the response
    return response.json()

# Define dynamic input parameters. Discover more about available parameters in the official documentation.
# Example:
input_parameters = {
    "sentence_level_timestamps": False,
    "diarization": False,
}

jobs = {}

# Send the transcription request
for uploaded_url in url_mappings.values():
    try:
        response = send_transcription_request(
            audio_url=uploaded_url,
            organization_name=ORGANIZATION_NAME,
            salad_key=SALAD_API_KEY,
            language_code="en",  # Specify the language code if needed
            input_params=input_parameters
        )
        print(f"Transcription request submitted successfully. Job ID: {response['id']}")

        jobs[uploaded_url] = response["id"]
    except Exception as e:
        print(f"Error sending transcription request: {e}")

# Save jobs to a JSON file
with open("jobs.json", "w") as f:
    json.dump(jobs, f, indent=4)
    print(f"Job IDs mapped to Salad Storage URLs saved to jobs.json")

Expected Output

If the request is successful, the API will return a job_id and our function will print it out:

Transcription request submitted successfully. Job ID: 6ea78b03-6ca4-4aeb-9136-4442fe8f49b2

You should also have a jobs.json file with the mappings:

{
  "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token": "1ea78b03-6ca4-4aeb-9136-4442fe8f49b1",
  "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token": "2ea78b03-6ca4-4aeb-9136-4442fe8f49b2",
  "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token": "3ea78b03-6ca4-4aeb-9136-4442fe8f49b3"
}

The job_id is a unique identifier for the transcription job. You can use it later to check the job status or retrieve the final result.

Step 6: Retrieve the Transcription Result

If you did not specify a webhook to receive the result, you can manually retrieve the transcription result using the job_id returned in Step 5.

Function to Retrieve Transcription Results

This function retrieves the result of a transcription job from the Salad Transcription API:

def get_transcription_result(job_id, salad_key, organization_name):
    """
    Retrieves the transcription result from the Salad Transcription API for a given job ID.

    Args:
        job_id (str): The unique identifier of the transcription job.
        salad_key (str): The Salad API key for authentication.
        organization_name (str): Name of the Salad organization.

    Returns:
        dict: The response from the API containing the job status and transcription result.
    """
    # Construct the Salad API endpoint for retrieving job results
    api_endpoint = f"https://api.salad.com/api/public/organizations/{organization_name}/inference-endpoints/transcribe/jobs/{job_id}"

    # Set headers
    headers = {
        "Salad-Api-Key": salad_key,
        "Content-Type": "application/json",
        "accept": "application/json",
    }

    # Send the request
    response = requests.get(api_endpoint, headers=headers)
    response.raise_for_status()

    # Extract and return the response
    return response.json()

for job_ids in jobs.values():
    # Retrieve the transcription result
    while True:
        # Retrieve the transcription result
        time.sleep(5) # Adjust this value if needed, depending on whether your transcription is long or short.
        try:
            result = get_transcription_result(
                job_id=job_ids,
                salad_key=SALAD_API_KEY,
                organization_name=ORGANIZATION_NAME
            )
            job_id=job_ids
            if result["status"] in ["created", "started", "running", "failed"]:
                print(f'Job ID: {job_id} status is {result["status"]}')
            elif result["output"]:
                with open(f"{job_id}.txt", "w") as f:
                    text = str(result["output"]["text"])
                    f.write(text)
                print(f'Job ID: {job_id} status is {result["status"]}')
                with open(f'output_{job_id}.json', 'w', encoding='utf-8') as f:
                    json.dump(result, f, ensure_ascii=False, indent=4)
                break
            else:
                print(f'Current job status is {result["status"]}')
        except Exception as e:
            print(f'Error retrieving transcription result: {e}')
            break

Expected Output

If the job is still processing:

Job ID: 1ea78b03-6ca4-4aeb-9136-4442fe8f49b1 status is running

If the job is completed:

Job ID: 1ea78b03-6ca4-4aeb-9136-4442fe8f49b1 status is succeeded

The completed transcriptions will be saved in a text file with the job_id as the filename:

 That's one small step for man, one giant leap for mankind. It's hard to imagine that nearly every consumer in America walks around with the total computational power of the Apollo 11 mission in their pocket. The technological advances of the last 50 years now afford us incredible processing power, an evolution in gaming, and nearly ubiquitous online connectivity. With the introduction of blockchain technologies, our youngest generations are also seeing the value in digital assets and the integration of the real and digital worlds. I watched the transition happen in real time. Back in the day, gaming was really single player. Even if you had people over, we were on the couch playing games on consoles together. That's where we left it. When we went home, the game didn't follow. But now there's not a moment where I couldn't pick up my computer or pick up a console and get someone on there with me. And even outside of that game lobby, if I've played a match of Starcraft 2 or Counter-Strike GO, it doesn't end there. We go and talk about it on Discord afterwards. The concept of digital ownership is very nebulous for certainly the baby boomer generation. myself as a millennial, it's hard to really comprehend and understand what does digital ownership mean. When you look at Gen Z who are true digital natives, they were brought up interacting with the internet, they spend their day to day socializing with their friends on the internet. When you bring along a new technology, the younger generation just gets it more quickly. They understand what blockchain technology is unlocking, which is the ability to assign value or property rights without relying on a centralized middleman or a centralized platform to arbitrate who owns what online. The entire identity of online personalities has changed. It started with a simple avatar on Facebook or Twitter to convey who you were. And now there's live streamers that are fully running motion capture suits and changing their physical identity, their voice, everything about them. And that is how they convey themselves to the world. I started developing mobile games when the App Store first launched. It's been more than a decade now. And seeing that transformation from small indie studios where it was one or two developers could make a game that became a worldwide phenomenon. It started off as a niche product. And today we see that companies are investing hundreds of millions of dollars into developing single titles. So that's been a change that I've witnessed in a relatively short amount of time. That trend is continuing today. The gaming space is now blending into what we would call the physical or reality space, where everything we touch has some sort of gamification feature to it. From making sure that you're sleeping the right amount of hours every night, to exercising on your Apple Watch. Gamification has perforated every part of our lives. A lot of times we get the question of why salad? It's a universal food. Every culture in the world has a salad. And it's only a salad when you mix the ingredients. And the individuality of the world right now allows people to keep their personal identity while contributing to a larger whole that's more meaningful. With the emergence of AI, there are so many exciting problems that we'll be taking on as a species. but the largest constraint is going to be compute resources. As AI comes online, its demand for compute is doubling every three and a half months. So to overcome this limiting factor, we're looking into every household around America and the globe to tap into the idle compute resources to really support this growing AI revolution. With Salad, we look at the growth in AI machine learning. It's exponential, and when we look at where can those needs be met, that's in the consumers' homes, on their desks. It's sitting there right now. That is the solution for the impending AI crisis that we have. Allowing users to participate in that, to monetize that resource that they have, is very important and the key part of why Salad exists. The more people who adopt into this system, not only will they be exposed to information and education about the systems you're interacting with, whether that's blockchain on the one hand, or if it's just gaming monetization models on the other hand. You learn more through interacting with the thing. Most people, especially for myself, I'm not exactly hands-on, but I need to be involved with something to start understanding it. It's kind of hard to just read what other people are saying, watch YouTube videos over and over again, and really get the idea. You need to get involved and get in it. As computers become more and more powerful, the data centers of the future will sit in everyone's home. Together, collectively, we'll build an infrastructure layer that powers different P2P networks, ecosystems, and economies. Together, we will build an infrastructure layer that will power our digital lives, maintain the integrity of the digital P2P networks, and Salad is here to empower that future. To learn more about how to make your computer work for you, visit salad.com. Salad. Share. Compute. For epic loot.

A full JSON response is also saved in a JSON file named output_JOB_ID.json for your further processing.

{
  "id": "ea78b03-6ca4-4aeb-9136-4442fe8f49b1",
  "input": {
    "sentence_level_timestamps": false,
    "diarization": false,
    "url": "https://storage-api.salad.com/organizations/YOUR_ORGANIZATION_NAME/files/audio/....mp3?token",
    "language_code": "en"
  },
  "inference_endpoint_name": "transcribe",
  "status": "succeeded",
  "events": [
    {
      "action": "created",
      "time": "2025-03-12T16:29:01.2808121+00:00"
    },
    {
      "action": "started",
      "time": "2025-03-12T16:29:01.4068381+00:00"
    },
    {
      "action": "succeeded",
      "time": "2025-03-12T16:29:37.9204505+00:00"
    }
  ],
  "organization_name": "salad",
  "output": {
    "text": " That's one small step for man, one giant leap for mankind. It's hard to imagine that nearly every consumer in America walks around with the total computational power of the Apollo 11 mission in their pocket. The technological advances of the last 50 years now afford us incredible processing power, an evolution in gaming, and nearly ubiquitous online connectivity. With the introduction of blockchain technologies, our youngest generations are also seeing the value in digital assets and the integration of the real and digital worlds. I watched the transition happen in real time. Back in the day, gaming was really single player. Even if you had people over, we were on the couch playing games on consoles together. That's where we left it. When we went home, the game didn't follow. But now there's not a moment where I couldn't pick up my computer or pick up a console and get someone on there with me. And even outside of that game lobby, if I've played a match of Starcraft 2 or Counter-Strike GO, it doesn't end there. We go and talk about it on Discord afterwards. The concept of digital ownership is very nebulous for certainly the baby boomer generation. myself as a millennial, it's hard to really comprehend and understand what does digital ownership mean. When you look at Gen Z who are true digital natives, they were brought up interacting with the internet, they spend their day to day socializing with their friends on the internet. When you bring along a new technology, the younger generation just gets it more quickly. They understand what blockchain technology is unlocking, which is the ability to assign value or property rights without relying on a centralized middleman or a centralized platform to arbitrate who owns what online. The entire identity of online personalities has changed. It started with a simple avatar on Facebook or Twitter to convey who you were. And now there's live streamers that are fully running motion capture suits and changing their physical identity, their voice, everything about them. And that is how they convey themselves to the world. I started developing mobile games when the App Store first launched. It's been more than a decade now. And seeing that transformation from small indie studios where it was one or two developers could make a game that became a worldwide phenomenon. It started off as a niche product. And today we see that companies are investing hundreds of millions of dollars into developing single titles. So that's been a change that I've witnessed in a relatively short amount of time. That trend is continuing today. The gaming space is now blending into what we would call the physical or reality space, where everything we touch has some sort of gamification feature to it. From making sure that you're sleeping the right amount of hours every night, to exercising on your Apple Watch. Gamification has perforated every part of our lives. A lot of times we get the question of why salad? It's a universal food. Every culture in the world has a salad. And it's only a salad when you mix the ingredients. And the individuality of the world right now allows people to keep their personal identity while contributing to a larger whole that's more meaningful. With the emergence of AI, there are so many exciting problems that we'll be taking on as a species. but the largest constraint is going to be compute resources. As AI comes online, its demand for compute is doubling every three and a half months. So to overcome this limiting factor, we're looking into every household around America and the globe to tap into the idle compute resources to really support this growing AI revolution. With Salad, we look at the growth in AI machine learning. It's exponential, and when we look at where can those needs be met, that's in the consumers' homes, on their desks. It's sitting there right now. That is the solution for the impending AI crisis that we have. Allowing users to participate in that, to monetize that resource that they have, is very important and the key part of why Salad exists. The more people who adopt into this system, not only will they be exposed to information and education about the systems you're interacting with, whether that's blockchain on the one hand, or if it's just gaming monetization models on the other hand. You learn more through interacting with the thing. Most people, especially for myself, I'm not exactly hands-on, but I need to be involved with something to start understanding it. It's kind of hard to just read what other people are saying, watch YouTube videos over and over again, and really get the idea. You need to get involved and get in it. As computers become more and more powerful, the data centers of the future will sit in everyone's home. Together, collectively, we'll build an infrastructure layer that powers different P2P networks, ecosystems, and economies. Together, we will build an infrastructure layer that will power our digital lives, maintain the integrity of the digital P2P networks, and Salad is here to empower that future. To learn more about how to make your computer work for you, visit salad.com. Salad. Share. Compute. For epic loot.",
    "duration_in_seconds": 325.488,
    "duration": 0.09,
    "processing_time": 35.118329763412476
  },
  "create_time": "2025-03-12T16:29:01.2808121+00:00",
  "update_time": "2025-03-12T16:29:37.9204505+00:00"
}

Full Script

Here is the full script combining all the steps:

# For extracting audio from Facebook
import yt_dlp

# For handling HTTP requests and file paths
import requests
from pathlib import Path
import time

# For working with JSON files (batch processing).
import json

def download_audio_from_Facebook(video_url, output_format="mp3"):
    """
    Downloads audio from a Facebook video in the specified format.

    Args:
        video_url (str): The URL of the Facebook video.
        output_format (str): Desired audio format (default: "mp3").

    Returns:
        str: The filename of the downloaded audio file.
    """
    try:
        ydl_opts = {
            'format': 'bestaudio/best',
            'postprocessors': [{
                'key': 'FFmpegExtractAudio',
                'preferredcodec': output_format,
                'preferredquality': '128',
            }],
            'outtmpl': '%(id)s.%(ext)s',
        }
        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
            info = ydl.extract_info(video_url, download=True)
            return f"{info['id']}.{output_format}"
    except Exception as e:
        print(f"Error downloading audio: {e}")
        return None


def upload_to_salad_storage(audio_file_path, salad_api_key, organization_name, mime_type="audio/mpeg"):
    """
    Uploads an audio file to Salad Storage and generates a signed access URL.

    Args:
        audio_file_path (str): Path to the audio file.
        salad_api_key (str): Salad API key for authentication.
        organization_name (str): Name of the Salad organization.
        mime_type (str): MIME type of the file (default: "audio/mpeg").

    Returns:
        str: The signed URL for the uploaded file.
    """
    try:
      # Generate the upload URL
      file_name = f"audio/{Path(audio_file_path).name}"
      upload_url = f"https://storage-api.salad.com/organizations/{organization_name}/files/{file_name}"

      # Prepare headers for authentication
      headers = {"Salad-Api-Key": salad_api_key}

      # Prepare data for the upload
      data = {
          "mimeType": mime_type,
          "sign": "true",  # Request a signed URL
          "signatureExp": 3 * 24 * 60 * 60  # 3 days in seconds
      }

      # Prepare the file to be uploaded
      with open(audio_file_path, "rb") as f:
          files = {"file": (Path(audio_file_path).name, f)}
          response = requests.put(upload_url, headers=headers, data=data, files=files)
          response.raise_for_status()

      # Extract and return the signed URL
      signed_url = response.json().get("url")
      if not signed_url:
          raise ValueError("Signed URL not returned by the API.")
      return signed_url

    except requests.exceptions.RequestException as e:
        print(f"Error uploading to Salad storage: {e}")
    except ValueError as ve:
        print(f"API response error: {ve}")
    return None


# Salad Storage credentials
SALAD_API_KEY = "YOUR_SALAD_API_KEY"  # Replace with your Salad API key
ORGANIZATION_NAME = "YOUR_ORGANIZATION_NAME"  # Replace with your organization name

# List of Facebook URLs
facebook_urls = [
    "https://www.facebook.com/user/videos/name-of-video/38947598374958739457/",
    "https://www.facebook.com/user/videos/name-of-video/39475903749587893745/",
    "https://www.facebook.com/user/videos/name-of-video/29837453749587394857/"
]

# Dictionary to store mappings of Facebook URLs to Salad Storage links
url_mappings = {}

for url in Facebook_URLS:
    try:
        print(f"Processing: {url}")
        audio_path = download_audio_from_Facebook(url)  # Step 1: Download audio
        print(f"Downloaded audio successfully, uploading to Salad Storage...")
        storage_url = upload_to_salad_storage(audio_path, SALAD_API_KEY, ORGANIZATION_NAME)  # Step 2: Upload audio
        url_mappings[url] = storage_url
        Path(audio_path).unlink()  # Clean up local file
        print(f"Processed successfully: {url}")
    except Exception as e:
        print(f"Failed to process {url}: {e}")

# Save mappings to a JSON file
with open("url_mappings.json", "w") as f:
    json.dump(url_mappings, f, indent=4)

print("Batch processing complete. URL mappings saved to url_mappings.json.")

def send_transcription_request(audio_url, organization_name, salad_key, language_code="en", input_params=None):
    """
    Sends a transcription request to the Salad Transcription API with dynamic input parameters.

    Args:
        audio_url (str): The signed Salad Storage URL for the audio file.
        api_endpoint (str): The Salad Transcription API endpoint.
        salad_key (str): The Salad API key for authentication.
        language_code (str): The language of the audio file (default: "en").
        input_params (dict): Additional input parameters to include in the request.

    Returns:
        dict: The response from the API, including the job ID.
    """
    # Set headers
    headers = {
        "Salad-Api-Key": salad_key,
        "Content-Type": "application/json",
        "accept": "application/json",
    }
    api_endpoint = f"https://api.salad.com/api/public/organizations/{organization_name}/inference-endpoints/transcribe/jobs"
    # Prepare the input dictionary, appending the audio URL and language code
    input_data = input_params.copy() if input_params else {}
    input_data["url"] = audio_url
    input_data["language_code"] = language_code

    # Request body
    data = {
        "input": input_data
    }

    # Send the request
    response = requests.post(api_endpoint, headers=headers, json=data)
    response.raise_for_status()  # Raise an error for bad responses

    # Extract and return the response
    return response.json()


def get_transcription_result(job_id, salad_key, organization_name):
    """
    Retrieves the transcription result from the Salad Transcription API for a given job ID.

    Args:
        job_id (str): The unique identifier of the transcription job.
        salad_key (str): The Salad API key for authentication.
        organization_name (str): Name of the Salad organization.

    Returns:
        dict: The response from the API containing the job status and transcription result.
    """
    # Construct the Salad API endpoint for retrieving job results
    api_endpoint = f"https://api.salad.com/api/public/organizations/{organization_name}/inference-endpoints/transcribe/jobs/{job_id}"

    # Set headers
    headers = {
        "Salad-Api-Key": salad_key,
        "Content-Type": "application/json",
        "accept": "application/json",
    }

    # Send the request
    response = requests.get(api_endpoint, headers=headers)
    response.raise_for_status()

    # Extract and return the response
    return response.json()


# Define dynamic input parameters. Discover more about available parameters in the official documentation.
# Example:
input_parameters = {
    "sentence_level_timestamps": False,
    "diarization": False,
}

jobs = {}

# Send the transcription request
for uploaded_url in url_mappings.values():
    try:
        response = send_transcription_request(
            audio_url=uploaded_url,
            organization_name=ORGANIZATION_NAME,
            salad_key=SALAD_API_KEY,
            language_code="en",  # Specify the language code if needed
            input_params=input_parameters
        )
        print(f"Transcription request submitted successfully. Job ID: {response['id']}")

        jobs[uploaded_url] = response["id"]
    except Exception as e:
        print(f"Error sending transcription request: {e}")

# Save jobs to a JSON file
with open("jobs.json", "w") as f:
    json.dump(jobs, f, indent=4)
    print(f"Job IDs mapped to Salad Storage URLs saved to jobs.json")

for job_ids in jobs.values():
    # Retrieve the transcription result
    while True:
        # Retrieve the transcription result
        time.sleep(5) # Adjust this value if needed, depending on whether your transcription is long or short.
        try:
            result = get_transcription_result(
                job_id=job_ids,
                salad_key=SALAD_API_KEY,
                organization_name=ORGANIZATION_NAME
            )
            job_id=job_ids
            if result["status"] in ["created", "started", "running", "failed"]:
                print(f'Job ID: {job_id} status is {result["status"]}')
            elif result["output"]:
                with open(f"{job_id}.txt", "w") as f:
                    text = str(result["output"]["text"])
                    f.write(text)
                print(f'Job ID: {job_id} status is {result["status"]}')
                with open(f'output_{job_id}.json', 'w', encoding='utf-8') as f:
                    json.dump(result, f, ensure_ascii=False, indent=4)
                break
            else:
                print(f'Current job status is {result["status"]}')
        except Exception as e:
            print(f'Error retrieving transcription result: {e}')
            break

​Prerequisites

​Step 1: Import Libraries

​Step 2: Define Functions

​Step 3: Define Variables

​Step 4: Batch Process Multiple URLs

​Expected Output

​Step 5: Send Prepared URLs to the Salad Transcription API

​Function to Send a Transcription Request

​Expected Output

​Step 6: Retrieve the Transcription Result

​Function to Retrieve Transcription Results

​Expected Output

​Full Script

Prerequisites

Step 1: Import Libraries

Step 2: Define Functions

Step 3: Define Variables

Step 4: Batch Process Multiple URLs

Expected Output

Step 5: Send Prepared URLs to the Salad Transcription API

Function to Send a Transcription Request

Expected Output

Step 6: Retrieve the Transcription Result

Function to Retrieve Transcription Results

Expected Output

Full Script