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

Nova-2 vs Chirp 3

2 models compared on the same benchmark — accuracy, latency, price, language coverage and capabilities. Chirp 3 leads on raw accuracy; Chirp 3 is fastest; Chirp 3 covers the most languages; Nova-2 is cheapest.

Open both in playground

Verdict · who wins what

Most accurate

Chirp 3

10.0% WER

Google Cloud

Fastest

Chirp 3

0.3× realtime

Google Cloud

Most languages

Chirp 3

68 languages

Google Cloud

Cheapest

Nova-2

$0.0058/min

Deepgram

Comparing

2 models · same benchmark

1

Batch & Realtime

Nova-2

Deepgram

$0.0058/min
#18 · 76.2

Deepgram's Nova-2 speech recognition

Use This ModelDetails →

2

Batch

Chirp 3

Google Cloud

$0.0107/min
#4 · 45.0

Google's latest generative ASR foundation model — 85+ languages

Use This ModelDetails →

Overview

30-day benchmark average

Overall score

0–100

76.2

45.0

Accuracy rank

of 35

#18

#4

Word error rate

WER

16.5%

10.0%

Character error rate

CER

11.9%

5.8%

Match error rate

MER

15.4%

9.7%

Word info lost

WIL

21.0%

14.1%

Languages

33

68

Cost vs. accuracy

overall WER vs price

Chirp 3

Nova-2

field · 33 models

lower-left is better

Accuracy by category

WER · lower is better

code_switching

63.7%
17.1%

conversational

10.2%
4.7%

finance

11.3%
7.1%

Accuracy by accent

WER · English

african

26.2%
20.0%

australian

15.4%
16.9%

indian

25.4%

Speed & latency

p50 → p99

Processing speed

× realtime

0.1×

0.3×

Latency p50

1.6s

11.0s

Latency p90

3.7s

11.2s

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

83.33%

85.71%

Error rate

16.67%

14.29%

Pricing

billed per second

Rate

per minute

$0.0058

$0.0107

Billing granularity

per 1s

per 15s

1 hour of audio

$0.35

$0.64

100 hours

Capabilities

Feature support

Speaker diarization

Yes

Yes

Word-level timestamps

Yes

No

Code-switching

Yes

Coverage & deployment

Languages · formats · regions

Languages

33

68

Mode

Batch & Realtime

Batch

Region

United States

United States

Audio formats

general

13.1%
8.2%

legal

3.8%
2.9%

medical

7.3%
1.9%

noisy

31.3%
29.2%

technical

4.9%
3.7%
27.0%

uk

11.9%
3.0%

us

9.4%
5.7%

Latency p99

3.7s

11.2s

Final drain

192ms

—

Real-time factor

1.02×

—

Flicker

10.5%

—

Cadence

0.8/s

—

$34.92

$64.08

1,000 hours

$349.20

$640.80

Yes

Auto-detect language

Yes

Yes

Live / streaming

Yes

No

Custom vocabulary

Yes

Yes

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

Max file size

2 GB

No limit

Max duration

No limit

8 h

Compliance

HIPAA
SOC2
GDPR
HIPAA
SOC2
GDPR

Frequently asked questions

Is Nova-2 or Chirp 3 more accurate?

Chirp 3 is more accurate, with a 10.0% word error rate versus 16.5% for Nova-2, on our standardized benchmark.

Which is cheaper, Nova-2 or Chirp 3?

Nova-2 is cheaper at $0.0058/min versus $0.0107/min for Chirp 3.

Should you choose Nova-2 or Chirp 3?

Choose Nova-2 for lower cost; choose Chirp 3 for accuracy, speed, and language coverage.