Cheapest Speech-to-Text APIs: Cost Comparison

This comparison covers 21 English base offers across 10 providers, using reference files of 10 minutes English audio at 0.01 GiB. At sampled monthly volumes of 600 and 6,000 minutes, Azure Speech MAI-2 Preview Promo is lowest. At sampled volumes of 21,600, 60,000, and 600,000 minutes, Speech is Cheap Subscription is lowest.

Workload Cost Scenarios

Pricing structures vary based on file length, billing increments, and feature add-ons. The following scenarios calculate costs across distinct engineering workloads using verified provider constraints.

Base English Transcription

Calculated scenario of 21,600 English minutes across 2,160 ten-minute files (0.01 GiB each) for base recorded transcription without add-ons.

Lowest calculated cost: Speech is Cheap Subscription at $20.00 ($0.000926/min)

Speech is Cheap Subscription yields the lowest calculated total cost, followed by Azure Speech MAI-2 Preview Promo and Speech is Cheap PAYG.

Show All Eligible Offers and Prices

Includes 21 maintained public pricing offers across 10 speech-to-text providers. Word timestamps and basic punctuation are included where bundled at no extra charge. Excludes Rev AI non-English rates because the workload is English, unverified diarization rates, and real-time streaming.

Base English Transcription
Provider OfferTotal CostEffective Cost per Actual Audio Minute
Speech is Cheap Subscription Speech is Cheap Pricing$20.00$0.000926
Azure Speech MAI-2 Preview Promo Promo Ends 2026-12-31 Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$36.00$0.001667
Speech is Cheap PAYG Speech is Cheap Pricing$43.20$0.002
Speechmatics Melia 1, Early Access Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$46.44$0.00215
AssemblyAI Universal-2 AssemblyAI Pricing$54.00$0.0025
Azure Speech Batch Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$64.80$0.003
Google Cloud Dynamic Batch, Chirp 3 Google Cloud Speech-to-Text Pricing ; Google Cloud Chirp 3 Features and Limits ; Google Cloud Recognition Configuration$64.80$0.003
OpenAI 4o Mini (Estimate, Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$64.80$0.003
Rev AI Reverb English Rev AI Pricing ; Rev AI Transcript Features ; Rev AI Changelog$72.00$0.003333
AssemblyAI Universal-3.5 Pro AssemblyAI Pricing$75.60$0.0035
Speechmatics Standard Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$86.40$0.004
Deepgram Nova-3 Monolingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$92.88$0.0043
OpenAI GPT-Transcribe OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$97.20$0.0045
Deepgram Nova-3 Multilingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$112.32$0.0052
Amazon Transcribe Standard Batch, US East Amazon Transcribe Pricing ; Amazon Transcribe US East Price List$129.60$0.006
Azure Speech Fast Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$129.60$0.006
OpenAI whisper-1 (Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$129.60$0.006
OpenAI 4o Diarize (Estimate, Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$129.60$0.006
Speechmatics Enhanced Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$144.00$0.006667
Gladia Recorded Starter Gladia Pricing ; Gladia Transcription Pricing Explained$219.60$0.010167
Google Cloud Standard, Chirp 3 Google Cloud Speech-to-Text Pricing ; Google Cloud Chirp 3 Features and Limits ; Google Cloud Recognition Configuration$345.60$0.016

Speaker Diarization on Short Files

Calculated scenario of 6,000 English minutes across 600 ten-minute files (0.01 GiB each) with speaker labels. Microsoft MAI2 preview documentation reports request failures around 15 minutes and longer when using diarization.

Lowest calculated cost: Azure Speech MAI-2 Preview Promo at $10.00 ($0.001667/min) Promo Ends 2026-12-31

Azure Speech MAI-2 Preview Promo yields the lowest calculated cost for speaker diarization on short files, followed by Speechmatics Melia 1 Early Access and AssemblyAI Universal-2.

Show All Eligible Offers and Prices

Limited to 18 offers with published pricing for speaker labels. Excludes Rev AI non-English rates. Azure Speech MAI-2 Preview Promo ends 2026-12-31, lacks an SLA, and Microsoft does not recommend it for production.

Speaker Diarization on Short Files
Provider OfferTotal CostEffective Cost per Actual Audio Minute
Azure Speech MAI-2 Preview Promo Promo Ends 2026-12-31 Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$10.00$0.001667
Speechmatics Melia 1, Early Access Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$12.90$0.00215
AssemblyAI Universal-2 AssemblyAI Pricing$17.00$0.002833
Azure Speech Batch Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$18.00$0.003
Google Cloud Dynamic Batch, Chirp 3 Google Cloud Speech-to-Text Pricing ; Google Cloud Chirp 3 Features and Limits ; Google Cloud Recognition Configuration$18.00$0.003
Rev AI Reverb English Rev AI Pricing ; Rev AI Transcript Features ; Rev AI Changelog$20.00$0.003333
AssemblyAI Universal-3.5 Pro AssemblyAI Pricing$23.00$0.003833
Speech is Cheap PAYG Speech is Cheap Pricing$24.00$0.004
Speechmatics Standard Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$24.00$0.004
Deepgram Nova-3 Monolingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$25.80$0.0043
Speech is Cheap Subscription Speech is Cheap Pricing$26.00$0.004333
Deepgram Nova-3 Multilingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$31.20$0.0052
Amazon Transcribe Standard Batch, US East Amazon Transcribe Pricing ; Amazon Transcribe US East Price List$36.00$0.006
Azure Speech Fast Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$36.00$0.006
OpenAI 4o Diarize (Estimate, Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$36.00$0.006
Speechmatics Enhanced Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$40.00$0.006667
Gladia Recorded Starter Gladia Pricing ; Gladia Transcription Pricing Explained$61.00$0.010167
Google Cloud Standard, Chirp 3 Google Cloud Speech-to-Text Pricing ; Google Cloud Chirp 3 Features and Limits ; Google Cloud Recognition Configuration$96.00$0.016

Word-Level Timestamps

Evaluates 6,000 minutes of recorded English speech across 600 ten-minute files (0.01 GiB each) with word timestamps enabled instead of speaker diarization. Suitable for alignment and subtitle generation.

Lowest calculated cost: Azure Speech MAI-2 Preview Promo at $10.00 ($0.001667/min) Promo Ends 2026-12-31

Azure Speech MAI-2 Preview Promo yields the lowest calculated cost, followed by Speechmatics Melia 1 Early Access and AssemblyAI Universal-2.

Show All Eligible Offers and Prices

Includes 16 offers with verified word timestamp pricing. Unverified or unpriced outputs are omitted.

Word-Level Timestamps
Provider OfferTotal CostEffective Cost per Actual Audio Minute
Azure Speech MAI-2 Preview Promo Promo Ends 2026-12-31 Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$10.00$0.001667
Speechmatics Melia 1, Early Access Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$12.90$0.00215
AssemblyAI Universal-2 AssemblyAI Pricing$15.00$0.0025
Azure Speech Batch Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$18.00$0.003
Speech is Cheap PAYG Speech is Cheap Pricing$18.00$0.003
Rev AI Reverb English Rev AI Pricing ; Rev AI Transcript Features ; Rev AI Changelog$20.00$0.003333
AssemblyAI Universal-3.5 Pro AssemblyAI Pricing$21.00$0.0035
Speech is Cheap Subscription Speech is Cheap Pricing$23.00$0.003833
Speechmatics Standard Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$24.00$0.004
Deepgram Nova-3 Monolingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$25.80$0.0043
Deepgram Nova-3 Multilingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$31.20$0.0052
Amazon Transcribe Standard Batch, US East Amazon Transcribe Pricing ; Amazon Transcribe US East Price List$36.00$0.006
Azure Speech Fast Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$36.00$0.006
OpenAI whisper-1 (Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$36.00$0.006
Speechmatics Enhanced Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$40.00$0.006667
Gladia Recorded Starter Gladia Pricing ; Gladia Transcription Pricing Explained$61.00$0.010167

Long-Form Audio Files

Evaluates 21,600 total minutes processed as 90 English files of 240 minutes and 0.5 GiB each without audio segmentation. Standard base transcription only without add-ons.

Lowest calculated cost: Speech is Cheap Subscription at $20.00 ($0.000926/min)

Speech is Cheap Subscription yields the lowest calculated cost for long-form files, followed by Speech is Cheap PAYG and AssemblyAI Universal-2.

Show All Eligible Offers and Prices

Restricted to nine shortlisted offers verified for these duration and size constraints: Speech is Cheap Subscription, Speech is Cheap PAYG, AssemblyAI Universal-2, Universal-3.5 Pro, Deepgram Nova-3 Monolingual PAYG, Nova-3 Multilingual PAYG, Rev AI Reverb English, AWS Standard Batch US East, and Azure Batch. Other offers are unranked because duration and size constraints were not verified. Exclusions include Gladia Starter (135-minute cap), Azure Fast/MAI (0.47 GiB cap), and Google Chirp 3 (which recommends 20 minutes as a non-hard limit). Speechmatics is not ranked because duration eligibility was not verified.

Spanish Language Transcription

Evaluates 21,600 minutes of Spanish-only recorded audio across 2,160 ten-minute files (0.01 GiB each). Measures calculated prices for a specific non-English workload without claiming universal language coverage.

Lowest calculated cost: Speech is Cheap Subscription at $20.00 ($0.000926/min)

Speech is Cheap Subscription yields the lowest calculated cost for Spanish files, followed by Azure Speech MAI-2 Preview Promo and Speech is Cheap PAYG.

Show All Eligible Offers and Prices

Restricted to an explicit three-offer shortlist across two providers: Speech is Cheap Subscription, Speech is Cheap PAYG, and Azure Speech MAI-2 Preview Promo. Other providers were not assessed for this Spanish example; this is a limited shortlist, not evidence of absent Spanish support and not a market-wide winner.

Spanish Language Transcription
Provider OfferTotal CostEffective Cost per Actual Audio Minute
Speech is Cheap Subscription Speech is Cheap Pricing$20.00$0.000926
Azure Speech MAI-2 Preview Promo Promo Ends 2026-12-31 Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$36.00$0.001667
Speech is Cheap PAYG Speech is Cheap Pricing$43.20$0.002

High-Volume Workload Scaling

Evaluates 600,000 minutes of recorded English audio across 60,000 ten-minute files (0.01 GiB each). Assesses public automatic volume tiers, but calculated costs do not prove concurrency capacity or quota availability.

Lowest calculated cost: Speech is Cheap Subscription at $555.60 ($0.000926/min)

Speech is Cheap Subscription yields the lowest calculated rate at scale, followed by Azure Speech MAI-2 Preview Promo and Speechmatics Melia 1 Early Access.

Show All Eligible Offers and Prices

Evaluates the base 21-offer set across 10 providers using published automatic pricing tiers. Excludes negotiated rates, custom enterprise contracts, and one-time promotional credits. Concurrency quotas must be verified separately.

High-Volume Workload Scaling
Provider OfferTotal CostEffective Cost per Actual Audio Minute
Speech is Cheap Subscription Speech is Cheap Pricing$555.60$0.000926
Azure Speech MAI-2 Preview Promo Promo Ends 2026-12-31 Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$1,000.00$0.001667
Speechmatics Melia 1, Early Access Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$1,044.90$0.001742
Speech is Cheap PAYG Speech is Cheap Pricing$1,200.00$0.002
AssemblyAI Universal-2 AssemblyAI Pricing$1,500.00$0.0025
Azure Speech Batch Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$1,800.00$0.003
Google Cloud Dynamic Batch, Chirp 3 Google Cloud Speech-to-Text Pricing ; Google Cloud Chirp 3 Features and Limits ; Google Cloud Recognition Configuration$1,800.00$0.003
OpenAI 4o Mini (Estimate, Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$1,800.00$0.003
Speechmatics Standard Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$1,944.00$0.00324
Rev AI Reverb English Rev AI Pricing ; Rev AI Transcript Features ; Rev AI Changelog$2,000.00$0.003333
AssemblyAI Universal-3.5 Pro AssemblyAI Pricing$2,100.00$0.0035
Deepgram Nova-3 Monolingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$2,580.00$0.0043
OpenAI GPT-Transcribe OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$2,700.00$0.0045
Deepgram Nova-3 Multilingual PAYG Deepgram Pricing ; Deepgram Recorded Audio API$3,120.00$0.0052
Speechmatics Enhanced Speechmatics Pricing and Volume Discounts ; Speechmatics Model Comparison$3,240.00$0.0054
Amazon Transcribe Standard Batch, US East Amazon Transcribe Pricing ; Amazon Transcribe US East Price List$3,600.00$0.006
Azure Speech Fast Azure Speech Pricing ; Azure Speech East US Retail Prices ; Microsoft MAI-Transcribe-2 Launch and Promotional Price ; Azure MAI-Transcribe Features and Preview Limits$3,600.00$0.006
OpenAI whisper-1 (Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$3,600.00$0.006
OpenAI 4o Diarize (Estimate, Sunset) OpenAI Whisper Model and Price ; OpenAI API Pricing ; OpenAI Diarization Model ; OpenAI File Transcription Guide$3,600.00$0.006
Gladia Recorded Starter Gladia Pricing ; Gladia Transcription Pricing Explained$6,100.00$0.010167
Google Cloud Standard, Chirp 3 Google Cloud Speech-to-Text Pricing ; Google Cloud Chirp 3 Features and Limits ; Google Cloud Recognition Configuration$9,000.00$0.015

Volume Cost Scaling

Transcription costs scale linearly for standard pay-as-you-go APIs but diverge when volume tiers or fixed subscriptions apply. Calculated expenses cover 600, 6,000, 21,600, 60,000, and 600,000 minutes across the base 21-offer set, displaying the lowest calculated offer with total cost and effective per-minute rates.

Volume Cost Scaling
Monthly Actual Audio MinutesLowest CostTotal CostEffective Cost per Actual Audio Minute
600Azure Speech MAI-2 Preview Promo$1.00$0.001667
6,000Azure Speech MAI-2 Preview Promo$10.00$0.001667
21,600Speech is Cheap Subscription$20.00$0.000926
60,000Speech is Cheap Subscription$55.56$0.000926
600,000Speech is Cheap Subscription$555.60$0.000926

Billing Increments on Short Audio Clips

Processing 6,000 six-second files equals 600 actual minutes of audio. Due to billing minimums, Speech is Cheap PAYG bills 6,000 minutes, while Rev AI Reverb English bills 1,500 minutes because Rev rounds each file up to a whole second with a 15-second minimum, not 15-second blocks.

Billing Increments on Short Audio Clips
ProviderBilled MinutesCalculated Cost
Speech is Cheap PAYG6,000$12.00 Speech is Cheap Pricing
Rev AI Reverb English1,500$5.00 Rev AI Pricing ; Rev AI Transcript Features ; Rev AI Changelog

Frequently Asked Questions

Which Speech-to-Text API Has the Lowest Calculated Cost?

There is no single lowest-cost provider across all scenarios; calculated costs depend on volume and feature requirements within our eligible comparison sets. Azure Speech MAI-2 Preview Promo wins the 600 and 6,000 minute base examples, as well as the 6,000 minute word timestamp and speaker examples. Speech is Cheap Subscription wins the 21,600, 60,000, and 600,000 minute base examples, along with the 21,600 minute long file scenario.

Why Are Certain Speech Providers Excluded from Long File Comparisons?

Long-form comparisons require processing 240-minute, 0.5 GiB audio files in a single request. Providers with payload caps (such as Azure Fast/MAI at 0.47 GiB), duration limits (Gladia Starter at 135 minutes), or unverified duration eligibility are omitted.

Are Introductory Credits and Negotiated Discounts Factored In?

Calculations reflect public automatic pricing tiers, standard pay-as-you-go rates, and published subscriptions. One-time onboarding credits, temporary trials, and privately negotiated enterprise discounts are excluded.

Sign Up for Speech is Cheap Subscription or PAYG

Choose Speech is Cheap Subscription for predictable monthly transcription volume or Speech is Cheap PAYG for flexible pay-as-you-go processing.

Speech API Billing Mechanics and Rules

All prices are in USD, with actual audio quantities in minutes. Speech is Cheap rounds files up to the next minute. Rev rounds up each file to the second with a 15-second minimum, not fixed billing increments.

Selected word or speaker features are billed according to each offer, though some include them. Speech is Cheap add-ons are charged on every rounded minute, including within the subscription allowance.

Calculations reflect direct commercial API charges. A cloud GPU is not mandatory to self-host open-source models; see the DIY hub to evaluate full infrastructure, hardware, and operational costs.

Comparisons apply eligibility filters based on published limits. Providers unable to verify support for single 240-minute, 0.5 GiB files are omitted from long-form comparisons. Unverified feature add-ons are excluded.

Certain rates depend on promotional terms. Azure Speech MAI-2 Preview Promo pricing ends on 2026-12-31; public preview has no SLA and Microsoft does not recommend it for production.

Public automatic tiers are included, calculating subscription fees, base allowances, overages, and add-ons together while excluding taxes, free credits, and negotiated rates. Introductory credits, recurring free allowances, and commitment discounts are distinct; see the master pricing table.

1 GiB = 1,073,741,824 bytes.

AssemblyAI File Size and Duration Limits ; Speech is Cheap Product Facts ; Speech is Cheap Product Facts ; Speech is Cheap Product Facts ; Rev AI Asynchronous API Overview ; Speech is Cheap Product Facts ; Google Cloud Chirp 3 Features and Limits ; Speech is Cheap Product Facts ; Amazon Transcribe Limits ; Speech is Cheap Product Facts ; Azure MAI-Transcribe Features and Preview Limits ; Azure Speech Quotas and Limits ; Speech is Cheap Product Facts ; Gladia File Formats and Limits ; Speech is Cheap Product Facts ; Speech is Cheap Product Facts Free Credits and Onboarding Discounts Guide