Best Video Transcription Tools in 2026

from deepgram import DeepgramClient, PrerecordedOptions

client = DeepgramClient(api_key="YOUR_KEY")

with open("meeting.mp4", "rb") as audio:
    payload = {"buffer": audio.read()}

options = PrerecordedOptions(
    model="nova-3",
    smart_format=True,
    diarize=True,
    topics=True,
    summarize="v2"
)

response = client.listen.rest.v("1").transcribe_file(payload, options)

for utterance in response.results.utterances:
    print(f"[Speaker {utterance.speaker}] {utterance.transcript}")
print(f"Summary: {response.results.summary.short}")

import whisper
import torch

# Load model (GPU recommended)
model = whisper.load_model("large-v3", device="cuda")

# Transcribe with language detection
result = model.transcribe(
    "interview.mp4",
    task="transcribe",
    word_timestamps=True,
    verbose=False
)

print(f"Detected language: {result['language']}")
for segment in result["segments"]:
    print(f"[{segment['start']:.1f}s - {segment['end']:.1f}s] "
          f"{segment['text']}")

import assemblyai as aai

aai.settings.api_key = "YOUR_KEY"

config = aai.TranscriptionConfig(
    speaker_labels=True,
    auto_highlights=True,
    content_safety=True,
    redact_pii=True,
    redact_pii_policies=[
        aai.PIIRedactionPolicy.person_name,
        aai.PIIRedactionPolicy.phone_number,
    ],
    summarization=True,
    summary_model=aai.SummarizationModel.informative,
    summary_type=aai.SummarizationType.bullets
)

transcriber = aai.Transcriber()
transcript = transcriber.transcribe("recording.mp4", config=config)

for utterance in transcript.utterances:
    print(f"Speaker {utterance.speaker}: {utterance.text}")
print(f"Summary: {transcript.summary}")

from google.cloud import speech_v2

client = speech_v2.SpeechClient()

config = speech_v2.RecognitionConfig(
    auto_decoding_config=speech_v2.AutoDetectDecodingConfig(),
    language_codes=["en-US", "es-US"],
    model="chirp_2",
    features=speech_v2.RecognitionFeatures(
        enable_automatic_punctuation=True,
        enable_word_time_offsets=True,
        diarization_config=speech_v2.SpeakerDiarizationConfig(
            min_speaker_count=2, max_speaker_count=6
        )
    )
)

with open("meeting.mp4", "rb") as f:
    content = f.read()

response = client.recognize(
    config=config, content=content,
    recognizer="projects/my-project/locations/global/recognizers/_"
)
for result in response.results:
    print(result.alternatives[0].transcript)

from rev_ai import apiclient

client = apiclient.RevAiAPIClient("YOUR_TOKEN")

# Submit for AI transcription with speaker diarization
job = client.submit_job_url(
    "https://storage/deposition.mp4",
    skip_diarization=False,
    language="en"
)

# Poll for completion
import time
while True:
    details = client.get_job_details(job.id)
    if details.status == "transcribed":
        break
    time.sleep(10)

# Get transcript with speaker labels
transcript = client.get_transcript_object(job.id)
for element in transcript.monologues:
    speaker = element.speaker
    text = " ".join([e.value for e in element.elements if e.type == "text"])
    print(f"Speaker {speaker}: {text}")

import speechmatics

# Real-time streaming transcription
ws = speechmatics.WebsocketClient(
    connection_url="wss://eu2.rt.speechmatics.com/v2",
    auth_token="YOUR_KEY"
)

config = speechmatics.TranscriptionConfig(
    language="en",
    enable_partials=True,
    operating_point="enhanced",
    diarization="speaker"
)

ws.add_event_handler(
    speechmatics.ServerMessageType.AddTranscript,
    lambda msg: print(f"[{msg['metadata']['start_time']:.1f}s] "
                      f"Speaker {msg.get('speaker', '?')}: "
                      f"{msg['metadata']['transcript']}")
)

with open("audio.wav", "rb") as f:
    ws.run_synchronously(f, config)

import boto3

transcribe = boto3.client("transcribe")

# Start a batch transcription job
transcribe.start_transcription_job(
    TranscriptionJobName="meeting-2026-01-15",
    Media={"MediaFileUri": "s3://recordings/meeting.mp4"},
    MediaFormat="mp4",
    LanguageCode="en-US",
    Settings={
        "ShowSpeakerLabels": True,
        "MaxSpeakerLabels": 6,
        "VocabularyName": "my-custom-vocab"
    },
    ContentRedaction={
        "RedactionType": "PII",
        "RedactionOutput": "redacted"
    }
)

# Get results (after polling for completion)
result = transcribe.get_transcription_job(
    TranscriptionJobName="meeting-2026-01-15"
)
print(result["TranscriptionJob"]["Transcript"]["TranscriptFileUri"])

# Otter.ai is primarily a SaaS product — API examples are for
# enterprise integrations. Most users interact via the app.

import requests

OTTER_API = "https://api.otter.ai/v1"
headers = {"Authorization": "Bearer YOUR_ENTERPRISE_TOKEN"}

# List recent transcriptions
speeches = requests.get(f"{OTTER_API}/speeches", headers=headers).json()

for speech in speeches["speeches"][:5]:
    print(f"{speech['title']} - {speech['created_at']}")
    print(f"  Duration: {speech['duration']}s")
    print(f"  Summary: {speech.get('summary', 'N/A')[:100]}")

import requests

GLADIA_API = "https://api.gladia.io/v2"
headers = {"x-gladia-key": "YOUR_KEY", "Content-Type": "application/json"}

# Submit transcription job
upload = requests.post(f"{GLADIA_API}/transcription", headers=headers, json={
    "audio_url": "https://storage/meeting.mp4",
    "diarization": True,
    "translation": True,
    "target_translation_language": "en",
    "code_switching": True
})
job = upload.json()

# Poll for results
import time
while True:
    result = requests.get(
        f"{GLADIA_API}/transcription/{job['id']}", headers=headers
    ).json()
    if result["status"] == "done":
        break
    time.sleep(5)

for utterance in result["result"]["transcription"]["utterances"]:
    print(f"[{utterance['speaker']}] {utterance['text']}")