Skip to content
Today
Sections

Build Reports disagree6 publishers 2 min read Published Updated

Gemini 3.8's cheap TTS tier still takes line-by-line stage directions

Google shipped two text-to-speech models with the same per-line performance control and put voice creation on only one of them, so the tier decision depends on whether a project has to invent a voice at all.

The Engineer · Build desk

Send a correction

Illustration accompanying Gemini 3.8's cheap TTS tier still takes line-by-line stage directions
Generated illustration

What happened

  • Google added two text-to-speech models to the Gemini family, Flash TTS for deep creative direction and character design and Flash-Lite TTS for high-volume, cost-efficient scale.
  • Only Flash TTS is named for generative voice design, building bespoke voices from scratch by prompting for role, accent and voice characteristics across more than 100 languages and dialects.
  • Voice replication recreates a vocal profile from a 30-second sample, with built-in consent verification, SynthID watermarking and C2PA credentials attached to the output.
  • Google places Flash TTS first and Flash-Lite second on Hume AI's Overall Quality Index, the only head-to-head ranking it cites for the cheaper model.

Why it matters

  • Who has to decide decision A dubbing queue or voice agent with a fixed cast can render on the cheaper model and keep its acting cues. Expressiveness is not what separates the tiers; the tier choice turns on whether the workload invents voices.
  • What it limits constraint Google did not publish a price for either model, so nobody outside can compute the monthly volume at which routing to Flash-Lite repays the engineering to split the pipeline.
  • Who's exposed exposure Thirty seconds is short enough that almost any recording a studio already holds is a usable sample. That shifts the legal risk on talent deals from the session booking to the consent verification step.
  • What's now possible capability Two-speaker staging from one script and hours of continuous audio with minimal drift remove the stitching and re-take passes that dialogue and audiobook pipelines currently budget for.

Both models take direction. Google's post says both TTS models give the user precise control over how each line is delivered [2], and it describes Flash-Lite with fine-grained control over tone, pacing and expressive nuance [4]. The performance section, which covers both, lists the same nonverbal tag syntax: `<laughs>`, `<sigh>`, `<gasp>`, plus active-listening interjections like `|mhm|` and `|yeah|` [6].

The catalogue sits on the Flash side. Google describes the voice studio as scaling up from 30 original voices to 2,000-plus production-ready voices, including regional varieties like Mexican Spanish, Quebec French and Scots English [19][20]. That is roughly 66 times the number of named voices [22].

Which raises the handoff question for anyone building a two-tier pipeline. Google says custom voices can be saved and managed to hold consistent performance with minimal drift across ongoing projects [5], but the post does not say whether a voice designed in the Flash studio can then be rendered by Flash-Lite. Without that, design-on-Flash and render-on-Flash-Lite is an assumption.

The quality evidence is third-party and thin on method. Google reports Flash TTS first overall on Hume AI's Voice Design Benchmark with 71.4, and leading accent modeling at 60.8 [10]. Blind human preference evaluations on Voice Arena put Flash and Flash-Lite in top positions in key global languages [11]. The 71.4 comes with no stated scale and no named field of competitors. It is a claim about someone else's prompt set. For it to transfer, the benchmark's briefs have to resemble yours: how many ask for an invented character, and how many just need a calm agent reading a refund policy.

The internal comparison is more legible. Google says the new models show major improvements on long-form content and dual-speaker screenplay control compared with Gemini 3.1 Flash TTS [12]. If you already run 3.1 Flash TTS on audiobook or dialogue work, you can measure that delta on your own scripts.

The two models follow 3.5 Live Translate, 3.5 Transcribe, 3.8 Live and 3.8 Live Extended Thinking in the Gemini Audio family [13], and Google says they improve user experiences in products including Gemini Notebook and Google Vids [7]. Voice remixing, which fine-tunes timbre, pitch, pace and accent from prompts such as "add subtle Southern US accent" or "soften the delivery", is listed as coming soon [8].

What to watch

  • Whether Voice Arena's public leaderboard reproduces the ordering Google reports once third parties run the same languages.
  • Whether SynthID detection for replicated-voice output is exposed to the people buying the audio, or only inside Google.
  • Whether the voice designer reaches end users in Google Vids and Gemini Notebook. That would change who the consent check is protecting.

Clarity's read

What the record supports and how the coverage leans. The claims behind it follow.

Reality

Evidence62
Adoption30
Hype gap+25
Incentives58
Confidence66

Perspective Coverage

7 publishers
Builder
Builder 64%
Operator
Operator 23%
Investor
Investor 13%
Why these scores

Claim ledger

Ranked by verification strength, evidence, and original report placement.

  1. [1]

    Google introduced two new text-to-speech models to the Gemini family: Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS.

    Reported Supported Source: blog.googleView cited source
  2. [2]

    Google states that both TTS models give the user precise control over how each line is delivered.

    Reported Supported Source: blog.googleView cited source
  3. [3]

    Voice replication recreates consistent vocal profiles from a 30-second audio sample of the user's voice or a voice they have the rights to use, backed by built-in consent verification, SynthID watermarking and C2PA credentials.

    Reported Supported Source: blog.googleView cited source
  4. [4]

    Gemini 3.8 Flash-Lite TTS is built for high-volume, cost-efficient scale, optimized for high-volume dubbing, audio content creation and expressive voice agents with fine-grained control over tone, pacing and expressive nuance.

    Reported Supported Source: blog.googleView cited source
  5. [5]

    Google says users can save and manage the custom voices they designed to ensure consistent performance and minimal drift across ongoing projects.

    Reported Supported Source: blog.googleView cited source
  6. [6]

    Scripted vocal bursts and backchanneling use nonverbal cues such as <laughs>, <sigh> and <gasp>, plus active-listening interjections such as |mhm| and |yeah|.

    Reported Supported Source: blog.googleView cited source
  7. [7]

    Google says the models enable improved user experiences in products like Gemini Notebook and Google Vids.

    Reported Supported Source: blog.googleView cited source
  8. [8]

    Voice remixing, letting users pick a library voice and fine-tune timbre, pitch, pace and accent with prompts such as "add subtle Southern US accent" or "soften the delivery", is listed as coming soon.

    Reported Supported Source: blog.googleView cited source
  9. [9]

    Native two-speaker scene staging directs multi-turn conversations from a single script while keeping both voices distinctly separated with natural conversational turn-taking.

    Reported Supported Source: blog.googleView cited source
  10. [10]

    Gemini 3.8 Flash TTS took the number one overall spot on Hume AI's Voice Design Benchmark with 71.4, and led in accent modeling with 60.8.

    Reported Supported Source: blog.google, citing Hume AI's Voice Design BenchmarkView cited source
  11. [11]

    In blind human preference evaluations on Voice Arena, Gemini 3.8 Flash and Flash-Lite TTS secured top positions amongst competitors in key global languages.

    Reported Supported Source: blog.google, citing Voice ArenaView cited source
  12. [12]

    Google says the models show major improvements on use cases such as long-form content and dual-speaker screenplay control compared to Gemini 3.1 Flash TTS.

    Reported Supported Source: blog.googleView cited source
  13. [13]

    The new TTS models follow 3.5 Live Translate, 3.5 Transcribe, 3.8 Live and 3.8 Live Extended Thinking in the Gemini Audio family.

    Reported Supported Source: blog.googleView cited source
  14. [14]

    Gemini 3.8 Flash TTS and Flash-Lite TTS took the number one and number two spots respectively on Hume AI's Overall Quality Index.

    Reported Supported Source: blog.google, citing Hume AI's Overall Quality IndexView cited source
  15. [15]

    Google's announcement post does not state a price for either Gemini 3.8 Flash TTS or Gemini 3.8 Flash-Lite TTS.

    Derived Supported
  16. [16]

    Google's announcement post does not state whether a custom voice designed with Gemini 3.8 Flash TTS can be rendered by Gemini 3.8 Flash-Lite TTS.

    Derived Supported
  17. [17]

    Gemini 3.8 Flash TTS is built for deep creative direction and character design, letting users create entirely new voices from scratch using natural language prompts and direct every performance line by line with granular control over acting cues, pacing, dialect shifts and backchanneling.

    Reported Contested Source: blog.googleView cited source
  18. [18]

    With Gemini 3.8 Flash TTS, users can create bespoke voices from scratch by customizing role, accent and voice characteristics across more than 100 languages and dialects using natural language prompting.

    Reported Contested Source: blog.googleView cited source
  19. [19]

    The Gemini voice library offers 2,000-plus production-ready voices with broad language coverage, including regional varieties such as Mexican Spanish, Quebec French and Scots English.

    Reported Contested Source: blog.googleView cited source
  20. [20]

    Google describes the Flash TTS vocal studio as scaling up from 30 original voices to an infinite library.

    Reported Contested Source: blog.googleView cited source
  21. [21]

    Long-form generation maintains voice quality, natural pacing and character timbre across hours of continuous audio with minimal speaker drift.

    Reported Contested Source: blog.googleView cited source
  22. [22]

    The 2,000-plus voice library is about 66 times the size of the 30-voice starting set.

    Derived Contested

Sources

6 independent publishers whose own reporting we read for this story.

  1. blog.google

    2 articles · September 23, 2026

    Gemini 3.8 text-to-speech says hello
  2. blog.vercel.com

    1 article · September 22, 2026

    Gemini 3.8 text-to-speech models now available on AI Gateway
  3. dev.to

    1 article · September 26, 2026

    [AI in Practice] Gemini 3.8 Flash TTS Launch: I built a "Learn Japanese with MVs" Web App and burned through my daily quota.
  4. runtimewire.com

    1 article · September 23, 2026

    Google launches Gemini 3.8 voice models with Hume founder Alan Cowen credited
  5. simonwillison.net

    2 articles · September 23, 2026

    Gemini 3.8 TTS Playground
  6. the-decoder.com

    2 articles · September 23, 2026

    Google's new Flash TTS models let you design AI voices from scratch using text descriptions
  7. thenewstack.io

    1 article · September 24, 2026

    OpenAI makes you call sales for a custom voice. Google just made it self-serve.

Topics and entities

Topics

Entities