Speech AI pricing

Compare speech AI pricing

Compare speech-to-text (STT), text-to-speech (TTS), and speech translation pricing across leading AI APIs. Enter your usage to compare costs between any two providers.

Pricing calculator

Stop overpaying for speech AI

1,000 hours of audio / month

1025501002505001k2.5k5k10k100k

Pricing assumptions

Based on public pay-as-you-go pricing. Enterprise discounts and committed-use contracts may differ. Some providers charge separately for certain features. The calculator uses the public price for the provider configuration that most closely matches Soniox.

Compare speech AI pricing on the same basis

Speech AI providers bill by minutes, hours, characters, tokens, models, and processing modes. We normalize public pricing into comparable workloads so you can see what STT, TTS, and speech translation actually cost at your usage level.

Public pricing: Based on published pay-as-you-go API rates.
Comparable configurations: Matched on the closest equivalent model, feature set, and processing mode.
Regularly reviewed: Rates, models, and assumptions are reviewed as providers update their APIs.

Compare speech-to-text and text-to-speech pricing

Compare speech-to-text, text-to-speech, and speech translation pricing across Soniox, OpenAI, Google, Azure, ElevenLabs, Deepgram, AssemblyAI, Speechmatics, Cartesia, and Fish Audio. Use the calculator to compare any two providers using the same workload.

Provider	Speech-to-Text	Text-to-Speech	Speech translation	Pricing updated
Soniox	Real-time, Async	Real-time	Real-time, Async	June 2026
OpenAI	Real-time, Async	Real-time	Real-time	June 2026
Google	Real-time, Async	Real-time	Real-time, Async	August 2026
Azure	Real-time, Async	Real-time	Real-time	June 2026
ElevenLabs	Real-time, Async	Real-time	Not listed	June 2026
Deepgram	Real-time, Async	Real-time	Not listed	May 2026
AssemblyAI	Real-time, Async	Not listed	Not listed	June 2026
Speechmatics	Real-time, Async	Real-time	Real-time, Async	June 2026
Cartesia	Real-time, Async	Real-time	Not listed	June 2026
Fish Audio	Not listed	Real-time	Not listed	August 2026

How speech-to-text pricing works

Speech-to-text (STT) APIs charge by audio duration, usually per minute or per hour, some providers bill by tokens. The rate depends on the model, whether transcription is real-time or asynchronous, and add-ons such as diarization.

Speech translation is usually priced on top of speech-to-text, either as a dedicated product or as separate transcription and translation steps.

How text-to-speech pricing works

Text-to-speech (TTS) APIs charge by the amount of text converted to audio, typically per character or per token, with rates that vary by model and voice tier.

Because providers bill in different units, the calculator converts TTS pricing into cost per hour of generated speech where the provider publishes enough to do so.

Compare speech AI costs at scale

A few cents per audio hour is easy to ignore in development. At production volume, the same difference shows up in your monthly infrastructure costs.

Compare your usage

Illustrative example

How much difference does $0.10 per audio hour bring to your monthly costs?

100 hours / month: $10 / month
1,000 hours / month: $100 / month
10,000 hours / month: $1,000 / month
100,000 hours / month: $10,000 / month

Illustrative example only, not tied to any provider. Use the calculator above to compare current costs for your usage.

Built for production speech AI

Soniox is one platform for real-time and asynchronous speech-to-text, speech translation, and text-to-speech. Compare Soniox pricing in the calculator above, then evaluate the API on your own workload.

Real-time and async speech-to-textTranscribe streaming and prerecorded audio across 60+ languages.
Integrated speech translationTranscribe and translate speech in the same real-time API call, at no extra cost.
Text-to-speechGenerate speech through the same developer platform.
Usage-based pricingPay only for the speech processing you use.

Build with Soniox View documentation

How we compare speech AI API pricing

Every provider in the calculator is priced by the same rules, so STT, TTS, and speech translation costs stay comparable even when billing units differ.

Public pay-as-you-go pricing: Every cost uses publicly listed API pricing. Private enterprise agreements and custom contracts are not the baseline.
Comparable configurations: We use the provider model or configuration that most closely matches the selected workload and processing mode, and say so in the price details.
Discounts and free tiers: Committed-use contracts, negotiated volume pricing, promotional credits, and free tiers are excluded. Costs are estimated from the first unit of usage.
Real-time vs asynchronous: When a provider charges different rates for streaming and batch speech-to-text, the calculator treats them as separate modes.
Feature-specific charges: Where a provider bills separately for a feature another provider includes, the add-on is listed in the price details next to the rate.
Text-to-speech normalization: Hourly TTS costs are derived from each provider’s published billing unit and the conversion shown in the price details.
Pricing updates: Every rate carries the month it was last reviewed. See the provider table or the calculator price details for each one.

Speech AI pricing FAQ

What is speech AI?

Speech AI is software that processes, understands, translates, or generates spoken language. It covers speech-to-text (STT), automatic speech recognition (ASR), text-to-speech (TTS), and speech translation.

What is the difference between speech AI and voice AI?

Speech AI refers to the underlying technologies that recognize, translate, or generate speech, such as STT and TTS. Voice AI usually means a full conversational system that combines speech processing with language models, business logic, and telephony. This page compares speech AI API pricing, not voice-agent platforms.

How much does speech-to-text cost?

Speech-to-text APIs charge by audio duration, usually per minute or per hour, and the rate depends on the provider, model, processing mode, and enabled features. Use the calculator above to compare STT costs by provider and monthly audio volume.

How much does text-to-speech cost?

Text-to-speech APIs charge by the amount of text converted to audio, usually per character or per token, with rates that vary by model and voice tier. The calculator converts TTS pricing to cost per hour of generated speech so providers can be compared on the same basis.

Does real-time transcription cost more than asynchronous transcription?

It depends on the provider. Some charge more for real-time streaming than for batch transcription, others price both the same. The calculator keeps the two modes separate whenever a provider publishes different rates.

Are enterprise discounts included?

No. Costs are based on public pay-as-you-go pricing. Negotiated enterprise contracts, committed-use discounts, and promotional credits are excluded.

How often is speech AI pricing updated?

Pricing is reviewed as providers change their models, APIs, and published rates. Each rate in the calculator and the provider table shows the month it was last checked. If you see a mismatched rate, please contact us so we can update it.

Spend less on speech AI infrastructure

Compare your workload, see the pricing differences, and choose the speech AI API that fits your application.

Compare pricing Build with Soniox