OpenTranscription
OpenTranscription
RankerModelsPlayground
OpenTranscription
OpenTranscription

One API to every speech-to-text model worth using. Compare them on your audio, route to the best one, pay per second.

Platform status

Product

RankerModelsTranscriptionsPlaygroundBlog

Developers

DocumentationReliabilityAPI VersioningStatus

Legal

Privacy PolicyTerms of ServiceSupport
© 2026 OpenTranscription
Back to Models

Amazon Transcribe vs Nova-2

2 models compared on the same benchmark — accuracy, latency, price, language coverage and capabilities. Amazon Transcribe leads on raw accuracy; Amazon Transcribe is fastest; Amazon Transcribe covers the most languages; Nova-2 is cheapest.

Open both in playground

Verdict · who wins what

Most accurate

Amazon Transcribe

13.0% WER

Amazon Web Services

Fastest

Amazon Transcribe

0.4× realtime

Amazon Web Services

Most languages

Amazon Transcribe

77 languages

Amazon Web Services

Cheapest

Nova-2

$0.0058/min

Deepgram

Comparing

2 models · same benchmark

1

Batch

Amazon Transcribe

Amazon Web Services

$0.0060/min
#8 · 51.5

AWS foundation model-powered ASR — 100+ languages

Use This ModelDetails →

2

Batch & Realtime

Nova-2

Deepgram

$0.0058/min
#18 · 76.2

Deepgram's Nova-2 speech recognition

Use This ModelDetails →

Overview

30-day benchmark average

Overall score

0–100

51.5

76.2

Accuracy rank

of 35

#8

#18

Word error rate

WER

13.0%

16.5%

Character error rate

CER

9.2%

11.9%

Match error rate

MER

12.1%

15.4%

Word info lost

WIL

16.7%

21.0%

Languages

77

33

Cost vs. accuracy

overall WER vs price

Nova-2

Amazon Transcribe

field · 33 models

lower-left is better

Accuracy by category

WER · lower is better

code_switching

55.8%
63.7%

conversational

6.9%
10.2%

finance

6.3%

Accuracy by accent

WER · English

african

24.6%
26.2%

australian

6.2%
15.4%

indian

3.2%

Speed & latency

p50 → p99

Processing speed

× realtime

0.4×

0.1×

Latency p50

12.4s

1.6s

Latency p90

13.0s

3.7s

Realtime / streaming

Streaming benchmark averages

Realtime rank

of 17

—

#4

Realtime score

0–100

—

78.0

Time to first word

—

986ms

Reliability

Rolling 30 days

Uptime

100.00%

83.33%

Error rate

0.00%

16.67%

Pricing

billed per second

Rate

per minute

$0.0060

$0.0058

Billing granularity

per 1s

per 1s

1 hour of audio

$0.36

$0.35

100 hours

$36.00

Capabilities

Feature support

Speaker diarization

Yes

Yes

Word-level timestamps

Yes

Yes

Code-switching

Yes

Coverage & deployment

Languages · formats · regions

Languages

77

33

Mode

Batch

Batch & Realtime

Region

United States

United States

Audio formats

11.3%

general

6.6%
13.1%

legal

6.2%
3.8%

medical

3.7%
7.3%

noisy

41.7%
31.3%

technical

3.2%
4.9%
25.4%

uk

0.0%
11.9%

us

1.9%
9.4%

Latency p99

13.0s

3.7s

Final drain

—

192ms

Real-time factor

—

1.02×

Flicker

—

10.5%

Cadence

—

0.8/s

$34.92

1,000 hours

$360.00

$349.20

Yes

Auto-detect language

Yes

Yes

Live / streaming

No

Yes

Custom vocabulary

No

Yes

mp3
wav
flac
m4a
ogg
webm
amr
mp3
wav
flac
m4a
ogg
webm

Max file size

2 GB

2 GB

Max duration

8 h

No limit

Compliance

HIPAA
SOC2
GDPR
HIPAA
SOC2
GDPR

Frequently asked questions

Is Amazon Transcribe or Nova-2 more accurate?

Amazon Transcribe is more accurate, with a 13.0% word error rate versus 16.5% for Nova-2, on our standardized benchmark.

Which is cheaper, Amazon Transcribe or Nova-2?

Nova-2 is cheaper at $0.0058/min versus $0.0060/min for Amazon Transcribe.

Should you choose Amazon Transcribe or Nova-2?

Choose Amazon Transcribe for accuracy, speed, and language coverage; choose Nova-2 for lower cost.