Google Cloud's Text-to-Speech
cloud.google.com · checked 15 Oct 2026
A Google Cloud service for producing speech audio from text or SSML across a broad voice and language catalog.
How it plugs in
Signal on Text-to-Speech Software7.8#13 of 223
Google Cloud Text-to-Speech is a paid API that produces spoken audio from text or SSML. Its catalog spans more than 380 voices in over 75 languages, with streaming, voice controls, custom models, and multiple audio formats. REST, gRPC, and client libraries connect it to applications on web, Android, and iOS. A free tier and trial credits are available.

Review
In a line
Text or SSML instructions can be submitted to Google Cloud Text-to-Speech to generate speech audio. It runs as a Google Cloud API and supports commercial use.
What it does
You send text or SSML and receive audio. SSML can control pauses, pronunciation, and how numbers or dates are spoken.
The catalog contains over 380 voices covering more than 75 languages and regional variants. Gemini-TTS handles one or multiple speakers and follows natural-language guidance for style, accent, pace, tone, and emotion.
Streaming synthesis is available for real-time conversations. Asynchronous long-audio jobs accept inputs of up to 1 million bytes. Pitch, speaking rate, and volume can also be adjusted.
How it connects
Applications can use REST or gRPC requests, documented APIs, authentication guidance, and official client libraries for languages such as Python, Java, Go, C#, Ruby, Node.js, and PHP. It also supports Application Default Credentials for programmatic sign-in.
Returned audio may use MP3, LINEAR16, OGG_OPUS, MULAW, or ALAW. Listed integrations include Cloud Vision API, YouTube, Unity, and Studio. The hosted service is proprietary, although its public API protocol definition uses the Apache 2.0 license.
It runs through API access on web, Android, and iOS applications.
Good fit
This suits teams adding spoken output to apps, devices, contact-center voicebots, or accessible electronic program guides. It also fits projects that need many languages, adjustable delivery, or streaming audio.
With as little as 10 seconds of audio, Chirp 3 instant custom voice can produce a personalized voice model for use in more than 30 locales.
Watch for
Billing counts input characters, including spaces, new lines, and SSML tags other than the mark tag. Billing must be enabled before using the service.
Check the voice, locale, output format, quotas, and byte limits for your workload. Instant custom voice has no free usage allowance, and Gemini TTS entries list no free usage limit.
Google describes security and compliance controls, but the service itself is not open source.
Pricing
Standard and WaveNet voices cost $4 a month, billed per 1 million characters after 4 million free characters. Neural2 costs $16 a month after 1 million free characters. Polyglot costs $16 a month after its 1 million-character free allowance.
Chirp 3 HD costs $30 a month after 1 million free characters; Studio costs $160 a month after 1 million free characters. Instant custom voice costs $60 a month per 1 million characters with no free usage.
Gemini 2.5 Flash TTS and Flash-Lite Preview TTS list $0.50 per 1 million text tokens and $10.00 per 1 million audio tokens. Gemini 3.1 Flash TTS lists $1.00 per 1 million text tokens and $20.00 per 1 million audio tokens. Gemini 2.5 Pro TTS lists the same token rates. No base price or free limit is listed for these Gemini entries.
New customers can receive $300 in Google Cloud credits to try Text-to-Speech and other Google Cloud products. A free plan and free trial are available.
Verdict
Choose this service if your product needs API-based speech, broad language coverage, adjustable voices, or streaming conversations. Its REST, gRPC, and client-library options give developers several ways to connect it, while free character allowances can cover smaller volumes.
Look elsewhere if you need open-source synthesis, a simple non-API editor, or predictable pricing without tracking characters or tokens.
Our take
Google Cloud Text-to-Speech fits applications seeking generated speech in many languages and audio formats. Streaming, SSML controls, and custom voices cover advanced use cases, while REST, gRPC, and client libraries support different application setups. Pricing varies by voice and counts characters or tokens, so estimate usage before committing. It is less suitable for buyers seeking open-source software or a standalone desktop tool rather than a cloud API.
- A catalog of over 380 voices covering more than 75 languages
- REST, gRPC, and official client libraries
- Streaming audio for real-time conversations
- MP3, LINEAR16, OGG_OPUS, MULAW, and ALAW output
- Free plan, free trial, and $300 in trial credits
- Billing counts characters, spaces, new lines, and most SSML tags
- Custom voice has no free usage allowance
- Hosted synthesis service is proprietary
- Pick it ifApps and devices that need multilingual spoken output; Contact-center voicebots and accessible program guides; Teams needing streaming or custom voices
- Skip it ifYou need open-source speech synthesis; You want a standalone editor instead of a cloud API; You need one simple price without character or token metering
Pricing
Instant custom voice is $60 a month with no free usage. Gemini TTS uses separate per-million-token rates, and no free limit is listed.
- Standard voicesper 1 million characters after 4 million free characters · 0–4 million characters free$4/mo
- WaveNet voicesper 1 million characters after 4 million free characters · 0–4 million characters free$4/mo
- Neural2 voicesper 1 million characters after 1 million free characters · 0–1 million characters free$16/mo
- Polyglot voicesPrice after free usage limit is reached · 1 million characters free usage$16/mo
- Chirp 3: HD voicesper 1 million characters after 1 million free characters · 0–1 million characters free · low-latency streaming · expressive conversational voices$30/mo
- Studio voicesper 1 million characters after 1 million free characters · 0–1 million characters free$160/mo
- Instant custom voiceper 1 million characters · no free usage · personalized voice models$60/mo
- Gemini 2.5 Flash TTS and Gemini 2.5 Flash-Lite Preview TTSInput: $0.50 per 1 million text tokens · Output: $10.00 per 1 million audio tokens · No free usage limit listedOn request
As cloud.google.com lists them, checked 15 Oct 2026.
Questions
How does Google Cloud Text-to-Speech work?
Text or SSML input is sent through the API, which returns spoken audio.
Which platforms can use it?
It provides API access for applications on web, Android, and iOS.
Does it offer a free tier?
Yes. A free plan and free trial are available. New customers can receive $300 in Google Cloud credits.
What audio formats are supported?
The service can return MP3, LINEAR16, OGG_OPUS, MULAW, and ALAW audio.
Can it create a custom voice?
Yes. Chirp 3 instant custom voice can create a personalized model from as little as 10 seconds of audio in more than 30 locales.
How is usage billed?
Character-based plans count input characters, including spaces, new lines, and most SSML tags. Gemini TTS plans use text and audio tokens.
Does it support live audio?
Yes. Streaming audio synthesis is available for low-latency, real-time conversations.