Google doesn't sell one text to speech product, it sells a menu, and the menu is priced in two different units. Some voices bill by the character and others by the token, so a buyer has to know which voice a call will use before the monthly figure means anything.
Google's strength is breadth, with 380+ voices across 75+ languages and variants on Google Cloud. A team that needs one voice to hold up on a phone call is choosing among Chirp 3: HD and Gemini-TTS, and those carry their own catches, starting with Preview labels and an instant voice cloning program that has closed to new requests.
What is Google text to speech, and how does it work?
Google text to speech is Google Cloud's Text-to-Speech API, one service that sells several families of voices, and as of November 2026 its newest are Chirp 3: HD voices and Gemini-TTS models, with Instant custom voice priced beside them. Older WaveNet, Standard, Neural2, Studio, and Polyglot voices stay on the price list as legacy models. Google says its Cloud Text-to-Speech API serves Gemini-TTS models named gemini-3.1-flash-tts-preview, gemini-2.5-flash-tts, gemini-2.5-flash-lite-preview-tts, and gemini-2.5-pro-tts, and says Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS aren't available through that API.
Streaming is bidirectional: a client sends text in and receives audio back as it is produced, and Google's examples use a client library's StreamingSynthesize call. Google describes its streaming as "ultra-low-latency speech" for real-time conversations, and Bland publishes 199ms to first audio.
What features does Google text to speech add?
Chirp 3: HD sets pace with speaking_rate from 0.25 to 2.0 and takes pause tags in a markup field, and Google says SSML, the XML markup older engines use for pauses and emphasis, isn't supported for its streaming requests. Gemini-TTS goes further. A style prompt drives the overall tone and delivery, and inline bracket tags add non-speech sounds such as [sigh] and [laughing], styles such as [whispering] and [sarcasm], and pauses such as [short pause]. Google also warns that vocalized-adjective tags such as [scared] are spoken aloud, which it calls likely an undesired side effect for most use cases. Gemini 3.1 Flash TTS handles single-speaker and multi-speaker dialogue.
Phone audio comes as MULAW and ALAW, the two telephony encodings, and Gemini-TTS streams PCM, ALAW, MULAW, or OGG_OPUS. Google has a text to speech service in Pipecat and a Google Cloud plugin for Python in LiveKit Agents.
How easy is Google text to speech to use?
Google text to speech is easy for a team already on Google Cloud and fiddly for one that isn't. Access comes through a Google Cloud project, and limits sit on the project rather than on a plan: Chirp 3: HD allows 200 requests per minute and 100 concurrent streaming sessions per project, and Gemini-TTS models carry per-minute limits such as 150 for gemini-2.5-flash-tts and 125 for gemini-2.5-pro-tts.
The harder part is the lineup. Three model generations sit side by side, and Google says the newest Gemini 3.8 models live on a different API than Cloud Text-to-Speech, so a team has to track which model sits behind which endpoint before building.
How much does Google text to speech cost?
Google text to speech costs $30 per million characters for Chirp 3: HD, with the first million characters free each month, and Gemini-TTS bills by the token instead of by the character. Google counts audio at 25 tokens per second of audio, so a minute of Gemini speech is 1,500 output tokens.
On Gemini 2.5 Flash TTS, Google charges $0.50 per million text tokens in and $10.00 per million audio tokens out, which works out to $0.015 per minute of audio on output alone. Gemini 2.5 Pro TTS and Gemini 3.1 Flash TTS charge $1.00 in and $20.00 out, or $0.03 per minute of audio on output, and Gemini-TTS has no free usage. Google prices Gemini 3.8 Flash TTS at $0.50 in and $9.00 out through 31 December 2026, then $1.00 and $18.00 from 1 January 2027. New Google Cloud customers get up to $300 in free credits.
What does Google text to speech cost next to Bland Speech?
Bland Speech costs $0.04 per 1,000 characters, which is $40 per million, with no monthly plan, and a new account gets 50,000 characters free once. A 60-second note costs $0.04. Enterprise pricing is custom, set with Bland's sales team.
Google is cheaper than Bland Speech on most of its menu. Chirp 3: HD, the legacy voices, and the Gemini models on output tokens all come in under $40 per million characters or $0.04 a minute, and only Studio voices and Instant custom voice cost more.
What these rows mean
- Bland Speech v3: Bland's standard rate, the same in the studio and the API, with no plan fee.
- Google character prices: Each is Google's rate after any free monthly usage, per million characters sent to the service.
- Google Gemini rows: Google bills tokens, not characters. Audio tokens are 25 per second of audio, so a minute of audio is 1,500 tokens and $0.015 or $0.03 of output cost, before the text input tokens.
Is Google text to speech secure enough for enterprise calls?
Google covers Text-to-Speech under Google Cloud's HIPAA business associate agreement, and says it enters into BAAs as necessary under HIPAA. Text-to-Speech also appears in Google Cloud's SOC 2 scope.
Bland is SOC 2 Type II audited, HIPAA compliant with a signed BAA, and FedRAMP 20x Class C certified at Moderate, granted on 17 November 2026, and Bland says speech synthesis runs inside that FedRAMP boundary. Bland runs its own model and speech, with zero AI sub-processors in the call path. Enterprise customers get dedicated orchestration and dedicated GPUs run by Bland.
Voice cloning is where Google's access story changes. Google says Chirp 3 Instant Custom Voice no longer takes new access requests and points customers to voice replication with Gemini 3.8 TTS, which Google labels Preview. Where Instant custom voice is available, Google requires a consent statement recorded in the chosen language and doesn't allow a custom script in its place.
Which sounds more human, Google text to speech or Bland Speech?
Bland Speech v3 ranked first of 13 AI voices in Design Arena's Audio Realism Bench results of 4 August 2026, at 1365, which was 256 points above Google's best-placed model, Gemini 2.5 Pro TTS, in fifth at 1109. Gemini 3.1 Flash TTS ranked eighth at 1037 and Gemini 2.5 Flash TTS tenth at 985, and a human recording scored 1500.
- #1BlandBland Speech v31365
- 2ndMicrosoftMAI-Voice-21224
- 3rdxAIGrok TTS1156
- 4thOpenAIGPT Realtime 21119
- 5thGoogleGemini 2.5 Pro TTS1109
Ranks are among 13 AI voices, and Google's own voice is on the list. Real human recordings scored 1500, at the right edge of this scale. Bars start at 600, not zero. Source: Design Arena's published results.
Chirp 3: HD, the legacy voices, and Gemini 3.8 weren't in those results, so Google's best-known voice names have no score from that test. Listeners hear two clips of the same line, without being told which model made which, and pick the one that sounds more like a real person, and the prompts include a Phone Agents category for customer support and routing calls.
Is Google text to speech worth it, or should a team choose Bland Speech?
Google text to speech is worth it for a team that wants low per-character prices on legacy and Chirp voices or one Google Cloud contract, and Bland Speech is the better choice for a phone line where callers judge the voice. A team comparing more vendors can read Google text to speech alternatives, and one moving off Google can follow how to migrate from Google text to speech.
Google text to speech pros and cons
Pros
- 380+ voices across 75+ languages and variants
- Gemini 2.5 Pro TTS ranked 5th of 13 AI voices on Audio Realism Bench on 4 August 2026
- First million Chirp 3: HD characters free each month, and legacy voices from $4 per million characters
- Style prompts and inline tags on Gemini-TTS for how a line is delivered
- MULAW and ALAW output for phone lines
- Text-to-Speech covered by Google Cloud's HIPAA BAA
Cons
- Two billing units, characters for most voices and tokens for Gemini-TTS
- Chirp 3 Instant Custom Voice has closed to new access requests, and Gemini 3.8 voice replication is in Preview
- Gemini-TTS has no free usage, and Google warns vocalized-adjective tags are spoken aloud
- Gemini 3.8 TTS models sit outside the Cloud Text-to-Speech API
- Limits are per Google Cloud project, such as 200 Chirp 3: HD requests per minute
Why do enterprises choose Bland Speech over Google text to speech?
Enterprises choose Bland Speech when callers judge a voice within the first second, when billing should run on one standard per-character rate, and when a security team would rather review one vendor than a cloud's whole catalog. Bland Speech v3 ranked first of 13 AI voices on Design Arena's listening test of 4 August 2026, and a team pays a standard $0.04 per 1,000 characters, directed or not.
Bland Speech's tags sit right before the words they shape, such as [laughs], [say warmly], [say in a whisper], or [chuckles], and Bland publishes the full set of emotion, intensity, and vocalization tags it says the model was trained on. Two controls from 0.0 to 1.0, expressiveness and stability, set how much intonation varies and how much the same input varies between renders, and a pause is written as <|N|> for 0.1 to 10.0 seconds. Output is mu-law at 8 kHz or PCM from 8 to 48 kHz, and the text to speech API page has the code.
An instant clone needs one clean sample of about 10 seconds, where Google's Instant custom voice has closed to new requests. A professional clone is fine-tuned on 30 minutes or more of audio after Bland's consent and rights review, and every clone is private to the customer's organization. Bland Speech also has a first-party Pipecat service and a LiveKit Agents plugin for Python, and the Bland Speech page has a studio to try it all.
Pros
- Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026
- One standard rate, $0.04 per 1,000 characters, with no tokens to convert
- An OpenAI-compatible endpoint, so an existing integration changes a base URL, a key, and a model name
- Performance tags for how a line is read
- Mu-law at 8 kHz and PCM from 8 to 48 kHz for phone lines
- Instant clones from about 10 seconds of audio and professional clones from 30 minutes or more
Cons
- More than a hobby project needs, because Bland Speech is built for production calls
Verdict: Google text to speech or Bland Speech for enterprise calls?
For enterprise phone calls where the voice should sound like a person, Bland Speech is the pick, with first place on Design Arena's 4 August 2026 results and a standard per-character rate. Google is the pick for a team that wants legacy voices at $4 per million characters or keeps every vendor inside one Google Cloud contract. The Bland Speech and Google text to speech comparison goes deeper, and the text to speech reviews collection has every vendor's review. The text to speech comparison hub links the rest.
Frequently asked questions
Is Google text to speech good for phone calls?
Google text to speech works for phone calls because it outputs MULAW and ALAW and streams audio as text arrives. Gemini 2.5 Pro TTS ranked 5th of 13 AI voices on Audio Realism Bench on 4 August 2026, against first for Bland Speech v3, and Chirp 3: HD wasn't in those results.
How much does Google text to speech cost?
Chirp 3: HD costs $30 per million characters after a free first million each month, legacy WaveNet and Standard voices cost $4, Neural2 costs $16, and Studio costs $160. Gemini-TTS bills tokens, from $10 per million audio tokens out on Gemini 2.5 Flash TTS. Bland Speech costs $0.04 per 1,000 characters.
What is the difference between Chirp 3: HD and Gemini-TTS?
Chirp 3: HD is billed by the character and controlled with speaking rate and pause tags, and Gemini-TTS is billed by the token and controlled with a style prompt and inline bracket tags. Google serves Gemini-TTS models through the same Text-to-Speech API, and puts Gemini 3.8 models on a different one.
Can Google text to speech clone a voice?
Google says Chirp 3 Instant Custom Voice no longer takes new access requests, and points to voice replication with Gemini 3.8 TTS, which is in Preview. Bland Speech makes an instant clone from about 10 seconds of audio and a professional clone from 30 minutes or more, and every clone is private to the customer's organization.
Is Google text to speech HIPAA compliant?
Google covers Text-to-Speech under Google Cloud's HIPAA BAA. Bland is HIPAA compliant with a signed BAA.
What is the best Google text to speech alternative?
Bland Speech is the best Google text to speech alternative for phone calls and voice agents, because Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026, and it bills a standard $0.04 per 1,000 characters. The text to speech API pricing comparison puts every rate side by side.
Sources
Google Cloud's Text-to-Speech pricing page and product page. Google Cloud's documentation on Chirp 3: HD, Gemini-TTS, streaming, Instant custom voice, SSML, output encodings, and quotas, plus its HIPAA and SOC 2 compliance pages. Pipecat's and LiveKit's text to speech pages. Bland's pricing and trust and security pages and Bland Speech documentation. Design Arena's published results.
