ElevenLabs vs Deepgram: transcription cost and accuracy

ElevenLabs Scribe v2 costs $0.22 per audio hour versus $0.258 for Deepgram Nova-3 and has the lower AA-WER (2.2% vs 5.2%); Nova-3 processes audio 8.0× faster (548.1× vs 68.4× real time).

Updated · 3 sources · Methodology

DimensionWinnerMarginEvidence
Price / audio hourElevenLabs$0.038$0.22 / $0.25823
AA-WER v2ElevenLabs3.0 pp2.2% / 5.2%1
Speed factorDeepgram8.0×68.4× / 548.1×1

Performance

ProviderModelAA-WER v2 (%)Speed factor
ElevenLabs1Scribe v22.2%68.4×
Deepgram1Nova-35.2%548.1×

Prices and request cost

ProviderPrice / audio hour100 h
ElevenLabs2$0.22$22.00
Deepgram3$0.258$25.80
Cost example: 100 hours of mono pre-recorded audio
ElevenLabs$22.00
Deepgram$25.80

Capabilities and serving conditions

DimensionElevenLabsDeepgram
Model and billing scopeScribe v2 is the file-transcription model; the separately priced realtime product is not the endpoint measured here.2Nova-3 offers pre-recorded and streaming APIs, but the displayed mono price and AA measurements describe pre-recorded processing.3

When each option fits

ElevenLabs

  • Choose Scribe v2 for transcripts that minimize correction work: it leads this pair on AA-WER and has the lower base audio price.123
  • At 68.4× versus Nova-3’s 548.1×, it is the slower choice for draining a large recording backlog.1

Deepgram

  • Choose Nova-3 when batch turnaround is the main constraint and the application can absorb a higher word-error rate.1
  • Nova-3’s 5.2% AA-WER exceeds Scribe v2’s 2.2%, and its mono base rate is also higher.123

Methodology and sources

AA-WER v2 uses AA-AgentTalk (50%), cleaned VoxPopuli (25%) and cleaned Earnings22 (25%). Speed factor is AA’s 7-day median of audio seconds processed per second on 10-minute audio, not realtime latency. Prices are base mono pre-recorded rates, excluding add-ons and extra channels.

Audio charges are estimates at listed rates, excluding tax and subscription credits. Novita AI publishes this speech API comparison using public sources.

  1. Artificial Analysis · AA-WER v2 non-streaming
  2. ElevenLabs · official API pricing
  3. Deepgram · official API pricing

Build with Novita Model APIs

Novita AI doesn’t offer speech-to-text. Pair your transcription provider with Novita LLM APIs to summarize and analyze transcripts, or Fish Audio text-to-speech for voice output.

Start BuildingExplore Models