How much does it cost to transcribe one hour of audio?
Transcribing one hour of recorded audio with a speech-to-text API costs $0.04 to $0.61, taking each of 12 vendors' cheapest model at list price: Groq's Whisper Large V3 Turbo is the cheapest and Gladia's Starter (async) the dearest. That is $0.0007 to $0.0102 a minute, and 6 of the 12 vendors charge $0.18 an hour or less. Having people type the same hour costs $119.40 with Rev AI's human transcription ($1.99 a minute), against $0.20 for its own Reverb (English) model.
Price your recording
1 hour of recorded audio: cheapest Groq Whisper Large V3 Turbo at $0.04.
| Option | This recording | Per minute |
|---|---|---|
| Groq Whisper Large V3 Turbo | $0.04 | $0.0007 |
| Groq Whisper Large V3 | $0.111 | $0.0019 |
| Speechmatics Melia 1 | $0.12 | $0.002 |
| AssemblyAI Universal-2 | $0.15 | $0.0025 |
| OpenAI gpt-4o-mini-transcribe Token-billed; per-minute figure is an estimate. | $0.18 | $0.003 |
| Google Cloud V2 dynamic batch | $0.18 | $0.003 |
| Azure AI Speech Batch transcription | $0.18 | $0.003 |
| Rev AI Reverb (English) | $0.20 | $0.0033 |
| AssemblyAI Universal-3.5 Pro | $0.21 | $0.0035 |
| ElevenLabs Scribe v2 | $0.22 | $0.0037 |
| Speechmatics Standard | $0.23 | $0.0038 |
| Deepgram Nova-3 Monolingual | $0.258 | $0.0043 |
| OpenAI gpt-transcribe | $0.27 | $0.0045 |
| Deepgram Whisper Large | $0.288 | $0.0048 |
| AssemblyAI Sync API | $0.30 | $0.005 |
| Rev AI Reverb (other languages) | $0.30 | $0.005 |
| Rev AI Whisper Large (English) | $0.30 | $0.005 |
| Gemini API Gemini 3.5 Transcribe Token-billed; per-minute figure is an estimate. | $0.306 | $0.0051 |
| Deepgram Nova-3 Multilingual | $0.312 | $0.0052 |
| OpenAI Whisper (whisper-1) | $0.36 | $0.006 |
| OpenAI gpt-4o-transcribe Token-billed; per-minute figure is an estimate. | $0.36 | $0.006 |
| OpenAI gpt-4o-transcribe-diarize Token-billed; per-minute figure is an estimate. | $0.36 | $0.006 |
| Amazon Transcribe Standard | $0.36 | $0.006 |
| Azure AI Speech Fast transcription | $0.36 | $0.006 |
| Azure AI Speech MAI Transcribe-1.5 | $0.36 | $0.006 |
| Azure AI Speech LLM Speech | $0.36 | $0.006 |
| Speechmatics Enhanced | $0.38 | $0.0063 |
| Gladia Starter (async) | $0.61 | $0.0102 |
| Google Cloud V2 standard | $0.96 | $0.016 |
| Google Cloud V1 with data logging | $0.96 | $0.016 |
| Google Cloud V1 without data logging | $1.44 | $0.024 |
| Rev AI Human transcription Typed by people. | $119.40 | $1.99 |
List prices for one recording, without free monthly minutes, one-off credits or promotions. Specialist medical and custom-trained models are left out; human transcription is shown for recorded files.
Every vendor at a minute, an hour and 10 hours
Each vendor's cheapest model for recorded audio:
| Vendor and model | 1 minute | 1 hour | 10 hours |
|---|---|---|---|
| Groq Whisper Large V3 Turbo | $0.0007 | $0.04 | $0.40 |
| Speechmatics Melia 1 | $0.002 | $0.12 | $1.20 |
| AssemblyAI Universal-2 | $0.0025 | $0.15 | $1.50 |
| OpenAI gpt-4o-mini-transcribe | $0.003 | $0.18 | $1.80 |
| Google Cloud V2 dynamic batch | $0.003 | $0.18 | $1.80 |
| Azure AI Speech Batch transcription | $0.003 | $0.18 | $1.80 |
| Rev AI Reverb (English) | $0.0033 | $0.20 | $2 |
| ElevenLabs Scribe v2 | $0.0037 | $0.22 | $2.20 |
| Deepgram Nova-3 Monolingual | $0.0043 | $0.258 | $2.58 |
| Gemini API Gemini 3.5 Transcribe | $0.0051 | $0.306 | $3.06 |
| Amazon Transcribe Standard | $0.006 | $0.36 | $3.60 |
| Gladia Starter (async) | $0.0102 | $0.61 | $6.10 |
| Rev AI Human transcription | $1.99 | $119.40 | $1,194 |
What can make a recording cost more
- Google Cloud: Each audio channel is billed separately.
- Amazon Transcribe: Up to two audio channels are included in the price.
- Deepgram: Stereo files are billed per channel: a 10-minute two-channel file counts as 20 minutes.
- AssemblyAI: Multichannel audio is billed per channel: a one-hour stereo file counts as two hours.
- AssemblyAI bills live audio for the time the connection is open, idle time included.
- Groq: Every request is billed for at least 10 seconds. That only matters for very short clips.
- Rev AI: Rounded up to the nearest second, with a 15-second minimum per file. That only matters for very short clips.
- Speaker labels cost extra on some APIs; the speech-to-text API pricing page lists each vendor's price.
Free monthly minutes can make a first hour cost nothing: Google Cloud includes 60 minutes a month and ElevenLabs includes 4 hours 30 minutes a month. Every free tier and credit is on free speech-to-text APIs.
Quick answers
How much does it cost to transcribe one hour of audio?
$0.04 to $0.61 with a speech-to-text API, each vendor's cheapest model at list price; the cheapest is Groq's Whisper Large V3 Turbo. Human transcription costs $119.40 an hour at Rev AI.
How much does transcription cost per minute?
$0.0007 to $0.0102 a minute of audio with an API, and $1.99 a minute for human transcription at Rev AI.
What does a 30-minute recording cost to transcribe?
$0.02 to $0.305 with an API, or $59.70 typed by people at Rev AI.
How much does it cost to transcribe 10 hours of audio?
$0.40 to $6.10 with an API at list prices, or $1,194 typed by people.
Does live transcription cost more?
Live (streaming) audio costs $0.15 to $1.02 an hour across the 10 vendors that price it, against $0.04 to $0.61 for recorded files. 9 of the 10 vendors that sell both charge more for live audio. The cheapest live option is AssemblyAI's Universal-Streaming.
Transcribing every month? The speech-to-text cost calculator prices monthly volume, and the cheapest speech-to-text API ranks the vendors.
Prices from each vendor's own pricing page, read on 6 October 2026 (ElevenLabs on 7 October 2026); the quoted text is on each vendor's page here. Amazon's prices are for US East (N. Virginia) and Azure's for East US.
Prices on this page were checked against the official pricing pages on .