· Updated 9 November 2026 · 01:06 CET
Gemini 3.8 Flash TTS: give the words a performance
Cast a voice, direct the delivery separately, and turn a script into narration or a two-person exchange with inline sounds and pauses.

A sentence can welcome someone, hold back a reveal, or sound as though it was written for a slide. The words may be identical; the delivery makes the difference. Gemini 3.8 Flash TTS is in the SUPERMODELS voiceover tool for shaping that performance.
Cast the voice and direct the read
Choose from 30 voices, put the spoken words in the script, and use the separate Style field for delivery. Keeping those parts apart helps make it clear what should be said and how it should sound. The model detects the language from the text and offers MP3 or WAV output.
Begin with a short passage and an empty Style field. Listen to the natural read before adding a direction such as “warm, measured, a little curious.” If you change the voice, keep the script fixed for the comparison. If you change the wording, keep the voice fixed. You will hear which decision is doing the work.
Write the exchange as two people
Choose a second voice to enable dialogue. Use exactly two names, with each turn on its own line. The first name is assigned the first voice; the next name receives the second. A direction in parentheses after a name applies to that turn.
Mira (quietly excited): I think the exhibition is ready.
Jon (curious): Even the room with the moving lights?
Mira: Especially that room.
This is a useful way to sketch a podcast opening, a character exchange or an explanation that benefits from a question and an answer. Keep each speaker’s role clear rather than giving both of them the same narration.
Put the pause where it belongs
Inline angle-bracket cues can add brief vocal sounds or pauses. For example, <chuckle> marks a moment in the script, while Style shapes the broader delivery. Use these cues sparingly, then listen for timing, pronunciation and whether the performance still sounds like the character you intended.
Open Gemini 3.8 Flash TTS, prepare a short script, and check the estimate before generating. A small audition is easier to revise than a finished narration.
Voice controls and script syntax checked against the official Gemini 3.8 Flash TTS documentation on Replicate.
Cover image prompt
Editorial cover artwork generated with OpenAI image generation. The exact prompt is below.
Illustrate this everyday object for the SUPERMODELS illustration collection: A small cobalt tabletop radio with a burnt-orange tuning dial on an ivory windowsill. A simple cable and shadow anchor the everyday object; the dial is blank without any lettering or numbers.
Contemporary SUPERMODELS editorial gouache illustration with clean cut-paper shapes, fine subtle risograph grain, warm ivory, cobalt blue, moss green and burnt orange, tactile matte painted paper, soft directional light, generous negative space and an elegant playful composition. Limit the palette to those colors and charcoal shadows. Use broad simplified shapes and visible dry gouache texture, as in a printed editorial illustration. Avoid photorealism, 3D rendering, glossy materials, ornate fantasy, clutter, outlines, gradients and collage grids. One self-contained full-bleed landscape artwork, exact 16:9 aspect ratio, with the focal subject comfortably inside the middle 75 percent. No text, lettering, logo, frame, borders or watermark.