Gemini 3.8 Flash TTS API

Zurelay charges $0.30 per 1M tokens of text in and $5.40 per 1M tokens of audio out. Google's list price is $0.50 and $9.00, so you save 40%. Gemini 3.8 Flash TTS counts 25 audio tokens for each second of speech, 1,500 a minute, so a minute costs about $0.008 plus the text you send. Google's price doubles on January 1, 2027, and Zurelay's follows it. Requests that fail cost nothing.

Create a Zurelay API key, set the SDK's base URL to https://api.zurelay.com/v1 and call client.audio.speech.create with model set to gemini-3.8-flash-tts, a voice and your text as input. Leave instructions out: this model reads everything it's given. The reply is the audio file itself: save it with write_to_file in Python, or write its bytes to a file in Node. To try voices first, use the Speech mode in the Playground at https://zurelay.com/app/playground.

Not on Zurelay. Google built it to read its input word for word, so a direction such as "Say cheerfully:" in the text would be spoken, and Google's own API takes a sustained style in a separate field that Zurelay doesn't pass on. A request with instructions is refused with an unsupported_parameter error instead of having them read out. Use vocal tags in the text for momentary sounds, or Gemini 3.1 Flash TTS for a style described in plain words.

Google documents angle-bracket tags for momentary sounds: <laugh>, <sigh>, <cough>, <breath> and <short pause>. Write them in input exactly where the sound should happen, such as "That's the best news all week. <laugh> Really." Zurelay sends your text to the model as written, tags included.

Gemini's 30 prebuilt voices, such as Kore, Puck, Charon, Leda, Aoede and Sulafat, with Kore as the default. OpenAI voice names work too and map to a Gemini voice of similar character: alloy speaks as Kore, nova as Leda, onyx as Orus and shimmer as Callirrhoe. Google's Extended Voice Library of more than 2,000 voices, voice design and voice replication aren't available through Zurelay.

More than 130, according to Google, against more than 70 for Gemini 3.1 Flash TTS. There's no language setting: the language is picked up from the text, so a Spanish sentence is spoken in Spanish.

Each request takes up to 5,000 characters of input; split longer text into several requests. Audio comes back as MP3 by default, or set response_format to wav, or to pcm for raw 24 kHz, 16-bit mono samples. Google lists an input limit of 8,192 tokens and an output limit of 16,384 for the model. Zurelay doesn't store the audio, so save the file you get back.

Not in one request on Zurelay. Google's own API can voice a multi-speaker dialogue in one call, but Zurelay speaks each request in a single voice. For a dialogue, make each line with its own voice and join the audio in order; raw PCM is the simplest to join.

A sentence or a short paragraph usually takes 6 to 9 seconds, and the whole file comes back in one reply. For a voice assistant, send a sentence at a time so playback starts sooner.

Yes. Google released it in preview in September 2026 (its Cloud documentation gives September 28) and has announced no shutdown date.

Yes. Your requests run on Google's Gemini 3.8 Flash TTS, and Zurelay doesn't swap in a different or smaller model. Speech is generated fresh on every call, so the same text and voice give a slightly different reading each time. Zurelay is an independent service and is not affiliated with Google.