Turn a finished script into expressive spoken audio with Google's Gemini 3.8 Flash TTS. Pick one of 70 preset voices, paste up to 5,000 characters, and generate a single-speaker voiceover for narration, explainers, and ads.
Creator planning guide
These are planning previews, not stored model outputs. Use them to choose a voice and tighten the script before running the live generator above.
Visual voice planning preview
Meet the new dashboard: a calmer way to review campaign performance, approve assets, and keep your team aligned before launch.
Visual voice planning preview
The storm arrived before sunset. Rain struck the windows, and somewhere beyond the garden wall, a bicycle bell rang once — then never again.
Visual voice planning preview
In this lesson we compare three approaches, then look at where each one fails in real production work.
When to use Flash
Flash is the fidelity-first Gemini 3.8 TTS model on AISnapEdit. Flash-Lite trades some range for speed and a lower rate.
Voices
70 presets
Grouped by use case so you can start from a voice built for your format.
Script length
5,000 characters
Each generation reads up to 5,000 characters as one continuous take.
Languages
130 (per Kie)
Kie lists 130 languages for Flash; write the script in the language you need.
Output
WAV audio
Kie documents the output as WAV, 16-bit PCM, 24 kHz, mono.
Long-form reads where a consistent, expressive voice carries the listener.
Scripts that need warmth and polish rather than the fastest possible draft.
Instructional audio where clear pacing and pronunciation matter.
Finish the script, pick a voice for the format, then review each take before generating the next section.
Write the exact words you want spoken. The model reads the text as written.
Start from the use-case group that matches your format, then compare one alternative.
Listen to the full take, then regenerate only the section that needs a change.
AISnapEdit exposes two inputs for Gemini 3.8 Flash TTS.
Text
Every generation
Paste the final script, up to 5,000 characters. Credits are charged per 1,000 characters, rounded up.
Voice
Tone and format
Pick one of 70 preset voices. Kore is the default; keep one voice per project for consistency.
Both Gemini 3.8 TTS models share the same 70 voices and inputs on AISnapEdit. Choose by what the audio is for.
Primary focus
Voice fidelity and expressive delivery.
Speed and cost efficiency.
Use Flash for audio your audience will listen to closely.
Languages (per Kie)
130 languages.
101 languages.
Check that your target language is covered before a long batch.
Rate on AISnapEdit
12 credits per 1,000 characters.
8 credits per 1,000 characters.
Flash-Lite keeps high-volume drafts cheaper.
Typical work
Narration, audiobooks, ads, and brand voiceovers.
Read-aloud features, bulk drafts, and short prompts.
Draft on Flash-Lite, finish on Flash if the delivery needs more range.
Each request reads the whole script in the selected voice. Dialogue between voices needs separate generations.
Generations use the 70 preset voices; there is no voice cloning or custom voice design on this page.
TTS cost is based on script length and shown before generation. If live pricing is unavailable, this page falls back to the recorded rate for this model.
credits per 1000 characters
Find answers to common questions about this model
It is Google's flagship Gemini 3.8 text-to-speech model for high-fidelity, expressive speech. On AISnapEdit it turns a script into single-speaker audio in one of 70 preset voices.
70 preset voices, grouped by use case such as tutor, commercial voiceover, storyteller, podcast, and professional. Kore is selected by default.
Up to 5,000 characters per generation. Split longer material into sections and keep the same voice for each part.
Cost is based on script length: 12 credits per 1,000 characters, rounded up to the next 1,000. The generator shows the cost before you create audio.
Kie documents Gemini 3.8 TTS output as WAV audio (16-bit PCM, 24 kHz, mono), which you can play and download from your results.
Choose Flash when voice quality and expressive delivery matter most. Choose Flash-Lite for faster, lower-cost drafts and high-volume read-aloud audio.
Paste your script, pick a voice, and check the credit estimate before you generate.
