- - 🗣️ Google's low-latency text-to-speech model for expressive speech.
- - 🆕 Adds inline audio tags to steer delivery and emotion.
- - 🎯 Steerable style prompts shape tone and pacing.
- - 📏 Generates audio with up to 32K token output context.
- - 🔒 Output watermarked with SynthID for AI-content identification.
- - 🏢 Available in public preview on Google AI Studio and Vertex AI.
- - 🧠 Based on the Gemini 3 Pro foundation model.
Google is an American multinational technology corporation and one of the world's most valuable brands. A subsidiary of parent company Alphabet Inc., Google operates across search, cloud computing, consumer electronics, and artificial intelligence. Its DeepMind and Google…
Explore 28 more models by Google →Gemini 3.1 Flash TTS is Google's text-to-speech model, released in April 2026 and offered in public preview through Google AI Studio and Vertex AI. It converts text into natural-sounding speech with low latency, letting developers steer delivery through descriptive style prompts and a set of inline audio tags that adjust tone, pacing, and emotion within an utterance. According to its DeepMind model card, the model is based on Gemini 3 Pro, produces audio with up to 32K token output, and watermarks all generated audio with SynthID.
Google positions the model for accessibility use cases such as screen readers and AAC devices, along with applications spanning audiobooks, gaming, and customer-facing voice systems. The expressive audio-tag system for narration control is among the headline features Google documents directly in its official materials.
Within Google's broader 2026 lineup, Gemini 3.1 Flash TTS handles the speech-synthesis modality, while text generation is served by siblings such as Gemini 3.5 Flash and Gemini 3.1 Pro Preview. It shares the Flash Audio designation with the real-time Gemini 3.1 Flash Live variant, covering the pre-generated portion of voice workflows.
Because no prior same-family TTS model appears in this catalog's sibling list, generational comparisons here are limited to the features Google describes in its official documentation: the inline audio-tag system and per-character style configuration.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usage · Gemini 3.1 Flash TTS · Organic sellers
| Seller | Health | Rep↓ | Sales | Routing | $ / 1M chars |
|---|---|---|---|---|---|
| ▲ Apex Ant 0x73b4…e736 | — | 81.71 | — | #2 | $93.75 |
| Venice.ai Proxy 0x1f22…18c9 | failingupstream error · 3.4 h | 86.84 | $4.49 | #1 | $103.125 |
"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.