Google TTS Chirp 3 HD

Specs & limits

• Google's latest text-to-speech generation, powered by its latest generative models and described as delivering realism and emotional resonance
• the comparison table names conversational agents as the intended use and lists it under Latest TTS models on the pricing page rather than Legacy. US$0.00003 per character (US$30 per 1 million characters) with the first 1 million characters free each month. Character counting (re-checked 2026-08-05): billable characters include spaces, newlines and all SSML tags except <mark>. Google's one-character-per-CJK-character note names only WaveNet and Standard voices and is SILENT on Chirp 3: HD - the guarantee is not stated for this tier

per Google official docs ↗

Use Google TTS Chirp 3 HD in 30 seconds

OpenAI-compatible: swap the base_url, keep your SDK. POST /v1/chat/completions

from openai import OpenAI

client = OpenAI(
    base_url="https://synthorai.io/v1",
    api_key="sk-syn-...",
)

resp = client.chat.completions.create(
    model="google-tts-chirp3-hd",
    messages=[{"role": "user", "content": "Summarize this diff"}],
)
print(resp.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://synthorai.io/v1",
  apiKey: "sk-syn-...",
});

const resp = await client.chat.completions.create({
  model: "google-tts-chirp3-hd",
  messages: [{ role: "user", content: "Summarize this diff" }],
});
console.log(resp.choices[0].message.content);

curl https://synthorai.io/v1/chat/completions \
  -H "Authorization: Bearer sk-syn-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google-tts-chirp3-hd",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

package main

import (
	"context"
	"fmt"

	"github.com/openai/openai-go/v3"
	"github.com/openai/openai-go/v3/option"
)

func main() {
	client := openai.NewClient(
		option.WithBaseURL("https://synthorai.io/v1"),
		option.WithAPIKey("sk-syn-..."),
	)
	resp, _ := client.Chat.Completions.New(context.TODO(), openai.ChatCompletionNewParams{
		Model: "google-tts-chirp3-hd",
		Messages: []openai.ChatCompletionMessageParamUnion{
			openai.UserMessage("Summarize this diff"),
		},
	})
	fmt.Println(resp.Choices[0].Message.Content)
}

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.chat.completions.*;

OpenAIClient client = OpenAIOkHttpClient.builder()
    .baseUrl("https://synthorai.io/v1")
    .apiKey("sk-syn-...")
    .build();

ChatCompletion resp = client.chat().completions().create(
    ChatCompletionCreateParams.builder()
        .model("google-tts-chirp3-hd")
        .addUserMessage("Summarize this diff")
        .build());
System.out.println(resp.choices().get(0).message().content().orElse(""));

About Google TTS Chirp 3 HD

• The comparison table names conversational agents as the intended use, and the voices page adds that these voices come in 30 distinct styles across many languages, suitable for real-time and standard applications, with advanced audio controls and low-latency real-time communication using text streaming.
• That last point is the sharpest line against Standard and Neural2: Chirp 3: HD is the only voice type in that table whose streaming column reads yes.
• Voices are named after stars rather than letters, Achernar, Algenib, Aoede, Charon, Kore, Leda, Puck, Sulafat, Zephyr, Zubenelgenubi and the rest, each published with a gender and combined with a locale into ids such as en-US-Chirp3-HD-Charon.
• Language coverage is a published BCP-47 list running from ar-XA and bn-IN through cmn-CN, ja-JP, and sw-KE to vi-VN, and the tier is generally available in the global, us, and eu endpoints plus asia-southeast1, europe-west2, and asia-northeast1.
• The default response format is LINEAR16; streaming adds ALAW, MULAW, OGG_OPUS, and PCM, while batch adds MP3 on top, so MP3 is not a streaming option here.
• Its controls are its own and all still Preview: pace through speaking_rate between 0.25 and 2.0, pause tags such as [pause short] and [pause long] that work only in the markup input field and never in text, custom pronunciations in IPA or X-SAMPA, and SSML restricted to synchronous requests with unlisted tags silently ignored.
• Google warns the model may disregard pause tags placed unnaturally.
• Pricing is US$0.00003 per character, quoted as US$30 per 1 million characters, with the first 1 million free, and the tier sits outside regionalization and data residency scope.
• Synthorai serves it over the OpenAI-compatible /v1/audio/speech endpoint.

FAQ

Related models

Compare