Google TTS Neural2

Specs & limits

The Neural2 voice tier of Google Cloud Text-to-Speech, still grouped under Legacy TTS models on the pricing page. US$0.000016 per character (US$16 per 1 million characters) with the first 1 million characters free each month, against Standard's 4 million. 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 Neural2 - the guarantee is not stated for this tier, so a CJK cost estimate from the per-character rate alone is not documented as reliable here.

per Google official docs ↗

Use Google TTS Neural2 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-neural2",
    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-neural2",
  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-neural2",
    "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-neural2",
		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-neural2")
        .addUserMessage("Summarize this diff")
        .build());
System.out.println(resp.choices().get(0).message().content().orElse(""));

About Google TTS Neural2

• The comparison table calls it general purpose, marks it generally available, and gives its controllability as SSML, the same profile Standard and WaveNet carry.
• Price is where the tier declares itself, at US$0.000016 per character, quoted as US$16 per 1 million characters, with the first 1 million characters each month free rather than Standard's 4 million; the pricing page still groups it under Legacy TTS models.
• The trade for that quality is reach.
• Where Standard spans locales from Afrikaans to Cantonese, Neural2 voice ids appear for a much shorter list, da-DK, de-DE, en-AU, en-GB, en-IN, en-US, es-ES, es-US, fr-CA, fr-FR, hi-IN, it-IT, ja-JP, ko-KR, pt-BR, th-TH, and vi-VN, named in the familiar locale-plus-letter form such as en-US-Neural2-F.
• Google notes Neural2 is available on global and single-region endpoints.
• Everything operational is shared with the rest of the classic API: SSML input, speakingRate from 0.25 to 2.0, pitch in semitones across a 20-point range either way, volumeGainDb from -96 to 16, an optional synthesis sample rate, the device profiles from wearables through telephony, and output as LINEAR16, MP3, OGG_OPUS, MULAW, or ALAW.
• The same 5,000-byte content ceiling per request applies, and like Standard, Neural2 is not offered over streaming synthesis, which is the concrete capability you give up against Chirp 3: HD.
• Billing counts characters including spaces and newlines, with all SSML tags except mark counted.
• Synthorai serves it over the OpenAI-compatible /v1/audio/speech endpoint.

FAQ

Related models

Compare