Summarizing virtual meetings

In this guide, we’ll show you how to use LLM Gateway to generate summaries of your virtual meetings. You’ll transcribe your audio and then use LLM Gateway to produce a custom summary.

The summarization, summary_model, and summary_type parameters on the transcription API are deprecated. Use LLM Gateway as shown below for more flexible and powerful summaries.

Get started

Before we begin, make sure you have an AssemblyAI account and an API key. You can sign up for a free account and get your API key from your dashboard.

Step-by-step instructions

Python
JavaScript

Create a new file and install the required packages.

pip install requests

Set up the API endpoint and headers.

import requests
import time

base_url = "https://api.assemblyai.com"

headers = {
    "authorization": "<YOUR_API_KEY>"
}

Create a new file and set up the API endpoint and headers.

import fs from "fs-extra";

const baseUrl = "https://api.assemblyai.com";

const headers = {
  authorization: "<YOUR_API_KEY>",
};

Python
JavaScript

Upload your local file to the AssemblyAI API and submit a transcription request.

with open("./my-audio.mp3", "rb") as f:
  response = requests.post(base_url + "/v2/upload",
                          headers=headers,
                          data=f)

upload_url = response.json()["upload_url"]

data = {
    "audio_url": upload_url,
    "speech_models": ["universal-3-pro", "universal-2"],
    "language_detection": True
}

response = requests.post(base_url + "/v2/transcript", json=data, headers=headers)

transcript_id = response.json()['id']
polling_endpoint = base_url + "/v2/transcript/" + transcript_id

while True:
    transcription_result = requests.get(polling_endpoint, headers=headers).json()
    if transcription_result['status'] == 'completed':
        break
    elif transcription_result['status'] == 'error':
        raise RuntimeError(f"Transcription failed: {transcription_result['error']}")
    else:
        time.sleep(3)

Upload your local file to the AssemblyAI API and submit a transcription request.

const path = "./my-audio.mp3";
const audioData = await fs.readFile(path);
let res = await fetch(`${baseUrl}/v2/upload`, {
  method: "POST",
  headers,
  body: audioData,
});
if (!res.ok) throw new Error(`Error: ${res.status}`);
const uploadResponse = await res.json();
const uploadUrl = uploadResponse.upload_url;

const data = {
  audio_url: uploadUrl,
  speech_models: ["universal-3-pro", "universal-2"],
  language_detection: true,
};

res = await fetch(`${baseUrl}/v2/transcript`, {
  method: "POST",
  headers: { ...headers, "Content-Type": "application/json" },
  body: JSON.stringify(data),
});
if (!res.ok) throw new Error(`Error: ${res.status}`);
const response = await res.json();

const transcriptId = response.id;
const pollingEndpoint = `${baseUrl}/v2/transcript/${transcriptId}`;

let transcriptionResult;
while (true) {
  res = await fetch(pollingEndpoint, { headers });
  if (!res.ok) throw new Error(`Error: ${res.status}`);
  transcriptionResult = await res.json();

  if (transcriptionResult.status === "completed") {
    break;
  } else if (transcriptionResult.status === "error") {
    throw new Error(`Transcription failed: ${transcriptionResult.error}`);
  } else {
    await new Promise((resolve) => setTimeout(resolve, 3000));
  }
}

Python
JavaScript

Send the transcript text to LLM Gateway with a summarization prompt.

prompt = """Provide a summary of this meeting transcript in bullet point format.
Focus on the key discussion points, decisions made, and any action items."""

llm_gateway_data = {
    "model": "claude-sonnet-4-6",
    "messages": [
        {"role": "user", "content": f"{prompt}\n\nTranscript: {transcription_result['text']}"}
    ],
    "max_tokens": 1500
}

response = requests.post(
    "https://llm-gateway.assemblyai.com/v1/chat/completions",
    headers=headers,
    json=llm_gateway_data
)

result = response.json()["choices"][0]["message"]["content"]
print(result)

Send the transcript text to LLM Gateway with a summarization prompt.

const prompt = `Provide a summary of this meeting transcript in bullet point format.
Focus on the key discussion points, decisions made, and any action items.`;

const llmGatewayData = {
  model: "claude-sonnet-4-6",
  messages: [
    {
      role: "user",
      content: `${prompt}\n\nTranscript: ${transcriptionResult.text}`,
    },
  ],
  max_tokens: 1500,
};

let res = await fetch("https://llm-gateway.assemblyai.com/v1/chat/completions", {
  method: "POST",
  headers: { ...headers, "Content-Type": "application/json" },
  body: JSON.stringify(llmGatewayData),
});
if (!res.ok) throw new Error(`Error: ${res.status}`);
const result = await res.json();

console.log(result.choices[0].message.content);

Understanding the response

The LLM Gateway returns the summary in the choices[0].message.content field. You can customize the format and style of the summary by adjusting the prompt.

Best practices

Customize the prompt to match your meeting type (e.g., standup, all-hands, client call).
For longer meetings, consider breaking the transcript into sections and summarizing each one separately.
Use Structured Outputs if you need the summary in a specific JSON format.
Try different LLM models to find the one that produces the best summaries for your use case.

Advanced usage

You can define custom summary formats by adjusting the prompt:

prompt = """Summarize this meeting transcript using the following format:
## Meeting Summary
- **Key Discussion Points**: [list main topics discussed]
- **Decisions Made**: [list any decisions]
- **Action Items**: [list action items with owners if mentioned]
- **Next Steps**: [list next steps]"""

Next steps

Summarization - Learn more about summarizing audio with LLM Gateway
LLM Gateway Overview - Explore all available models and features
Structured Outputs - Constrain responses to a specific JSON schema

Documentation Index

​Get started

​Step-by-step instructions

​Understanding the response

​Best practices

​Advanced usage

​Next steps

Get started

Step-by-step instructions

Understanding the response

Best practices

Advanced usage

Next steps