Google: Gemini 3.8 Flash Lite TTS

Google: Gemini 3.8 Flash Lite TTS is a text to speech model from Google. It belongs to the Gemini 3.8 TTS family, where it sits as the fast, high throughput member alongside Gemini 3.8 Flash TTS. Its main category is speech, and it was released by the provider on 2026-09-23.

The model takes text as its only input and produces speech as its only output. It has a context window of 32768 tokens and a maximum output of 16384 tokens. Thirty voices are available: Zephyr, Puck, Charon, Kore, Fenrir, Leda, Orus, Aoede, Callirrhoe, Autonoe, Enceladus, Iapetus, Umbriel, Algieba, Despina, Erinome, Algenib, Rasalgethi, Laomedeia, Achernar, Alnilam, Schedar, Gacrux, Pulcherrima, Achird, Zubenelgenubi, Vindemiatrix, Sadachbia, Sadaltager and Sulafat. The voice list is the main lever you have over the output, so picking a voice that matches the tone of your content is part of the integration work.

Because the input is text and the output is audio, the practical uses follow from turning written material into spoken material. You can generate narration for articles, documentation or long form content, since the 32768 token context window lets you pass substantial passages in a single request. The 16384 token maximum output bounds how much speech one call can produce, so very long documents are better split into sections. The range of voices makes the model usable for multi speaker style projects, such as dialogue between distinct characters, or for producing several variants of the same script so you can compare how different voices read the same lines. Applications that read content aloud, produce audio versions of written material, or generate voice assets for downstream editing all fit the text in, speech out shape of this model.

You call it through api14's OpenAI compatible API using the model id "google/gemini-3.8-flash-lite-tts". The model supports no request parameters beyond what the endpoint itself accepts, so there are no extra tuning options to set on the request. Send your text as the input and select a voice from the list above; the response contains the generated speech.

If your task needs something other than speech output, this model is not the right fit, since speech is its only output modality. It also accepts only text as input, so audio or image inputs are out of scope. For text generation or other modalities, look at other models on api14.