Google Cloud Text-to-Speech
cloud.google.com
- Free plan
- Web
- iPhone and iPad
- Android
- API
What Google Cloud Text-to-Speech is
Google Cloud Text-to-Speech is an API that converts text or SSML into audio data for applications. It offers more than 380 voices across more than 75 languages and variants, with requests available through REST or gRPC. Output formats include MP3, LINEAR16, OGG_OPUS, MULAW, and ALAW. Gemini-TTS supports single- or multispeaker output and lets users direct style, accent, pace, tone, and emotional expression with natural-language prompts. Chirp 3 instant custom voice can create a personalized voice model from as little as 10 seconds of audio and is available in more than 30 locales. The service supports streaming synthesis and long-audio input up to 1 million bytes. Controls include pitch, speaking rate, and volume gain. Google cites contact-center voicebots, device voice generation, and accessible electronic program guides as use cases. Billing must be enabled, and input characters count toward charges, including spaces, newlines, and SSML tags other than the mark tag. Standard and WaveNet voices have 0 to 4 million characters free, then cost US$4 per 1 million characters. New customers can receive $300 in Google Cloud credits to try this and other Google Cloud products.
Who it suits
It suits developers building applications that need generated speech through an API, including voicebots, device voices, or accessible program guides. Teams needing multilingual choices, speech controls, or streaming synthesis may also find the listed capabilities relevant.
Good to know, and what to watch
- More than 380 voices across 75-plus languages.
- REST and gRPC API access.
- Supports streaming audio synthesis.
- Several output formats are available.
- Billing must be enabled to use the service.
- Input characters, spaces, and tags count toward billing.
- Free usage limits vary by voice model.
Plans
Every plan and priceStandard voices
WaveNet voices
Neural2 voices
Polyglot voices
Getting started
- Open the Google Cloud Text-to-Speech website
- Choose REST or gRPC integration
- Provide text or SSML input
- Select a voice, language, or custom voice option
- Choose an MP3, LINEAR16, OGG_OPUS, MULAW, or ALAW output
- Enable billing before making requests
- Use the free trial or listed free character allowances where available
Limits to check
Standard and WaveNet voices include 0 to 4 million characters free, while Neural2 and Polyglot include 1 million characters free. Chirp 3 HD and Studio include 1 million characters free; Instant custom voice has no free usage limit, and Gemini plans list no free usage limit.
The Club's review
Google Cloud Text-to-Speech is a cloud API that turns text or SSML into speech audio. It best suits developers adding spoken interaction to apps, devices, or customer services. Its wide voice range and detailed speech controls make it flexible, while model-specific pricing calls for careful cost planning.
Overview
More than 380 voices across over 75 languages and variants give teams room to serve varied audiences without relying on a single voice style. REST and gRPC integration make this a service to build into software, rather than a ready-made reading app for someone who simply wants text spoken aloud.
Streaming synthesis supports low-latency conversations, and long-audio synthesis accepts up to 1 million bytes of input. Billing must be enabled. Charges count input characters, including spaces, newlines, and SSML tags other than the mark tag, so formatting can affect the bill as well as model choice.
Key features
Voice direction and personalization
Gemini-TTS can create single- or multispeaker speech and follow natural-language prompts for style, accent, pace, tone, and emotional expression. That suits applications where delivery needs to respond to context, though its token-based pricing differs from the per-character rates used by other voice options.
Chirp 3 instant custom voice can create a personalized voice model from as little as 10 seconds of audio and supports more than 30 locales. This is useful for products that need a distinctive voice, but its separate charge and lack of a free usage allowance make experimentation less forgiving than with some other models.
Control, integration, and output
Developers can adjust pitch by up to 20 semitones, speaking rate to as much as four times faster or slower than normal, and volume gain from -96 dB to +16 dB. Applications can receive MP3, Linear16, OGG Opus, MULAW, or ALAW audio. These controls and formats offer room to adapt speech to an application, but using them requires API integration.
Google identifies contact-center voicebots, voice generation in devices, and accessible electronic program guides as use cases. Its documentation includes guides, API references, quotas and limits, release notes, and a support resource. Google Cloud also maps security, privacy, and compliance controls to standards and undergoes independent verification through certifications, attestations, and audit reports.
Pricing
The service is paid, with a starting price of $4. New customers can receive $300 in Google Cloud credits to try Text-to-Speech and other Google Cloud products. Listed monthly plan amounts and usage charges vary by voice type:
| Voice option | Monthly plan | Usage and free allowance |
|---|---|---|
| Standard | Free | US$0.000004 per character (US$4 per 1 million characters); 0 to 4 million characters free |
| WaveNet | Free | US$0.000004 per character (US$4 per 1 million characters); 0 to 4 million characters free |
| Neural2 | $16 a month | Price after free usage limit is reached; 1 million characters free usage |
| Polyglot | $16 a month | Price after free usage limit is reached; 1 million characters free usage |
| Chirp 3: HD | Free | US$0.00003 per character (US$30 per 1 million characters); 0 to 1 million characters free |
| Studio | Free | US$0.00016 per character (US$160 per 1 million characters); 0 to 1 million characters free |
| Instant custom voice | Free | US$0.00006 per character (US$60 per 1 million characters); no free usage limit |
Standard and WaveNet offer the clearest low-cost starting point, with up to 4 million characters free and the same stated per-character rate. Chirp 3: HD and Studio also have a free character allowance, but their rates are higher. Instant custom voice has no free usage limit, which matters if you plan to assess it through substantial generation.
Gemini 2.5 Flash TTS and Gemini 2.5 Flash-Lite Preview TTS are priced at $0.50 per 1 million text tokens and $10.00 per 1 million audio tokens, with no free usage limit listed. Gemini 3.1 Flash TTS (Preview) is priced at $1.00 per 1 million text tokens and $20.00 per 1 million audio tokens, with no free usage limit listed. Gemini 2.5 Pro TTS has the same stated token rates as Gemini 3.1 Flash TTS. Gemini 3.8 Flash TTS (Preview), through Dec 31, 2026, is priced at $0.50 per 1 million text tokens and $9.00 per 1 million audio tokens, with no free usage limit. These token-based options need a different usage estimate from character-priced voices.
Platforms
Google Cloud Text-to-Speech is available through API and web platforms. API access, REST and gRPC requests, and commercial use are supported. Voice cloning is also supported.
Who it's for
This is a strong fit for development teams building voice into customer interactions, connected devices, or accessibility features. The range of languages, streaming support, speech controls, and custom voice option serve products with specific delivery needs. It is a weaker fit for individuals seeking a standalone text reader, or for teams that need predictable costs without first choosing a model and estimating character or token use.
Pros and cons
- Pro: More than 380 voices across over 75 languages and variants support products serving varied audiences.
- Pro: Streaming, speech controls, and single- or multispeaker Gemini-TTS give developers options for responsive and directed speech.
- Pro: Multiple output formats and REST or gRPC integration suit applications with different technical requirements.
- Con: API-centered access takes development work, making it less suitable as an out-of-the-box reading tool.
- Con: Rates differ by voice model and billing unit, while some options have no free usage allowance; cost estimates require selecting a model and accounting for input.
- Con: Billing counts spaces, newlines, and most SSML tags, so formatted input can raise usage charges.
Alternatives
For a free open-weight option, Kokoro TTS offers an Apache-licensed 82-million-parameter model. Browse Text-to-Speech Software for more options in the category, or see AI Voice Generators and Game Voice Generators for related directories.
Other alternatives include Altered Studio, AnnoMotion, LOVO AI, Adobe Firefly, Neocortex, Speechless, and Algebras AI Game Localization Agent.
Verdict
Choose Google Cloud Text-to-Speech if you are building an application that needs broad voice coverage, streaming, or fine control over speech. Its main advantage is the range of ways to shape and deliver audio through an API. Look elsewhere if you want a standalone reader or need a single, simple rate: model-specific pricing and input-counting rules make budgeting part of the implementation.
The verdict
Google Cloud Text-to-Speech offers a broad voice selection and controls through API requests, with multiple synthesis modes. Check the per-model usage terms and character-counting rules before estimating costs.
Questions people ask
What does Google Cloud Text-to-Speech cost?
Pricing is paid from $4. Standard and WaveNet voices are billed at US$4 per 1 million characters after their free allowance; other models have separate rates.
Which output formats are available?
The listed formats are MP3, LINEAR16, OGG_OPUS, MULAW, and ALAW.
Does it support custom voices?
Yes. Chirp 3 instant custom voice can create a personalized model from as little as 10 seconds of audio and is available in more than 30 locales.
Is there a free plan or trial?
A free plan and free trial are listed. New Google Cloud customers can receive $300 in credits to try Text-to-Speech and other Google Cloud products.
How is usage billed?
Billing counts input characters, including spaces, newlines, and SSML tags except the mark tag. Billing must be enabled to use the service.
What the maker publishes
- API access
- Yes source
- API positioning
- The product is an API powered by Google AI technologies for generating natural-sounding speech. source
- Audio formats
- The service can return formats including MP3, Linear16 and OGG Opus. source
- Authentication
- Client libraries support Application Default Credentials for programmatic authentication. source
- Billing limit
- Billing counts input characters including spaces, newlines, and SSML tags except the mark tag, and billing must be enabled to use the service. source
- Chirp 3 HD
- Chirp 3 HD voices provide low-latency streaming, expressive intonation and human disfluencies for conversational agents. source
- Commercial use
- Yes source
- Company founded
- Google's official history dates the company's beginnings to 1998. source
- Company headquarters
- Google's official history identifies the Googleplex in Mountain View, California as its headquarters. source
- Company legal name
- Google's terms identify the provider as Google LLC. source
- Company size
- LinkedIn lists Google as a public software-development company with 10,001-plus employees and headquarters in Mountain View, California. source
- Compliance
- Google Cloud describes ISO 27001/27017/27018/27701, SOC 1/2/3, PCI DSS and FedRAMP offerings plus GDPR and HIPAA alignment. source
- Custom voices
- Chirp 3 instant custom voice can create personalized voice models with as little as 10 seconds of audio input and is available in more than 30 locales. source
- Developer APIs
- Google documents REST v1 and v1beta1 APIs, gRPC APIs, authentication guidance and client libraries. source
- Documentation depth
- Documentation includes product basics, quickstarts, SSML tutorials, voice catalogs, API references, authentication, samples and release notes. source
- Export formats
- MP3,LINEAR16,OGG_OPUS,MULAW,ALAW source
- Free plan
- No source
- Gemini-TTS
- Gemini-TTS supports single- or multispeaker speech and lets users steer style, accent, pace, tone, and emotional expression with natural-language prompts. source
- Gemini-TTS
- Gemini-TTS supports single- and multispeaker synthesis with natural-language control of style, accent, pace, tone and emotion. source
- GetApp feature ratings
- GetApp reports AI voices 4.8/5, custom voices 5.0/5, multilingual support 5.0/5, API 5.0/5 and SSML 5.0/5 from small review subsets. source
- Instant custom voice
- Instant custom voice can create a personalized voice model from as little as 10 seconds of audio in more than 30 locales. source
- Integrations
- GetApp lists integrations including Cloud Vision API, YouTube, Unity and Studio. source
- Interfaces and formats
- Applications can integrate through REST or gRPC requests, and output formats include MP3, Linear16, and OGG Opus. source
- Latest release
- On December 30, 2025, Chirp 3 HD added preview synthesis for Chinese Hong Kong (yue-HK). source
Checked against the maker's pages on 15 Oct 2026.