Turn meetings into transcripts with speaker labels

The problem

A plain transcript can merge several people into one voice. An agent reading it may assign a decision or promise to the wrong person. A text model alone cannot recover speaker boundaries from an audio file it has never processed.

How NativePort helps

NativePort connects your recording to speech recognition that can separate speaker turns and return timestamps. Your agent receives a transcript organized by voice, giving it better material for meeting notes. Speaker labels distinguish voices; you add the actual names when you know them.

Technical implementation

Use Python 3 with requests installed (python -m pip install requests). Set NATIVEPORT_API_KEY in your environment to your NativePort key. Run the snippets on your server, where your key stays private.

Set AUDIO_PATH to the meeting’s MP3 recording. Request diarization and combine consecutive word tokens from the same speaker.

python

import os
import requests

BASE = "https://api.nativeport.ai"
HEADERS = {"Authorization": f"Bearer {os.environ['NATIVEPORT_API_KEY']}"}

with open(os.environ["AUDIO_PATH"], "rb") as audio:
    response = requests.post(BASE + "/elevenlabs/v1/speech-to-text",
        headers=HEADERS,
        data={"model_id": "scribe_v1", "diarize": "true"},
        files={"file": ("recording.mp3", audio, "audio/mpeg")},
        timeout=180)
response.raise_for_status()
transcript = response.json()
turns = []
for word in transcript.get("words", []):
    speaker = word.get("speaker_id") or "unknown"
    if not turns or turns[-1]["speaker"] != speaker:
        turns.append({"speaker": speaker, "start": word.get("start"), "text": ""})
    turns[-1]["text"] += word["text"]
for turn in turns:
    print(f"[{turn['start']}] {turn['speaker']}: {turn['text'].strip()}")

Retain the timestamps so a reviewer can jump back to the recording. Overlapping speech can still need correction before you ask an agent to attribute commitments. See diarization options.

Tool costs

Tool used in the example	Price per call
ElevenLabs — speech to text with speaker labels	$0.026375 / call

Prices in USD. Usage-based tools have no fixed per-call price. View pricing.

Let your agent build it

You don’t need to write this code yourself. Copy this page’s link and paste it into your agent. Ask it to follow the guide and implement the feature for you.

https://nativeport.ai/use-cases/meeting-transcripts/

The problem

How NativePort helps

Technical implementation

Tool costs

Let your agent build it

Keep exploring

Add spoken replies to Telegram and Discord bots

Analyze sentiment and review spoken content

Let your agent read spoken messages

Explore all use cases