テキストを音声に変換 エージェント 音楽 音声からテキストへ 音声クローン Voice Design Marketplace モア

TTSアリーナ

公式ベンチマーク、コミュニティ評価、サイドバイサイド比較。

無料登録

サイドバイサイド比較

テキストを入力し、2 つのモデルを選び、結果を比較します。フリーティアモデルにはアカウントが必要ありません。

フリーモデルはアカウントなしで動作する。 登録 高級車を比較するのに使われます

モデルリーダーボード

# モデル オフィシャル 共同体 あなたの評価 スピード 動物
1
Kokoro
Kokoro
Lightweight 82M parameter model delivering studio-quality speech with blazing-fast inference.
82M 1200h 2024
4.8 /5 5.0 /5
1 投票
fast Free
2
CosyVoice 2
CosyVoice 2
Alibaba's scalable streaming TTS with human-parity naturalness and near-zero latency.
300M 200000h 2024
4.26 /5 まだ投票はありません
medium Standard
3
Chatterbox
Chatterbox
State-of-the-art zero-shot voice cloning with emotion control from Resemble AI.
300M 2025
4.25 /5 まだ投票はありません
medium Premium
4
StyleTTS 2
StyleTTS 2
Human-level text-to-speech through style diffusion and adversarial training.
100M 585h 2024
4.23 /5 まだ投票はありません
medium Premium
5
Piper
Piper
A fast, local neural text to speech system optimized for Raspberry Pi and embedded devices.
15M 2023
4.15 /5 まだ投票はありません
fast Free
6
MeloTTS
MeloTTS
High-quality multilingual text-to-speech that runs on CPU with minimal latency.
25M 2024
4.13 /5 まだ投票はありません
fast Free
7
Dia TTS
Dia TTS
Multi-speaker dialog generation model that creates natural conversations between speakers.
1.6B 2024
4.09 /5 まだ投票はありません
medium Standard
8
VITS
VITS
Conditional variational autoencoder with adversarial learning for end-to-end text-to-speech.
25M 585h 2021
4.0 /5 まだ投票はありません
fast Free
9
Orpheus
Orpheus
Human-level emotional TTS model trained on 100K hours of speech data.
3B 100000h 2025
4.0 /5 まだ投票はありません
medium Standard
10
OpenVoice
OpenVoice
Instant voice cloning with granular control over style, emotion, and accent.
300M 2024
4.0 /5 まだ投票はありません
medium Premium
11
IndexTTS-2
IndexTTS-2
Zero-shot TTS with fine-grained emotion control and high expressiveness.
300M 2025
3.91 /5 まだ投票はありません
medium Standard
12
Spark TTS
Spark TTS
Voice cloning TTS with controllable emotion and speaking style via prompts.
500M 2025
3.9 /5 まだ投票はありません
medium Standard
13
Parler TTS
Parler TTS
Describe the voice you want in natural language and Parler generates matching speech.
880M 45000h 2024
3.83 /5 まだ投票はありません
medium Standard
14
Tortoise TTS
Tortoise TTS
Multi-voice text-to-speech focused on quality with autoregressive architecture.
400M 50000h 2022
3.7 /5 まだ投票はありません
slow Premium
15
Bark
Bark
Transformer-based text-to-audio model that generates realistic speech, music, and sound effects.
350M 100000h 2023
3.57 /5 まだ投票はありません
slow Standard
16
Bark Small
Bark Small
Lighter version of Bark with faster inference and lower memory usage.
150M 100000h 2023
— まだ投票はありません
medium Standard
17
GPT-SoVITS
GPT-SoVITS
Few-shot voice cloning TTS that replicates any voice from just 5 seconds of audio.
200M 2024
— まだ投票はありません
slow Standard
18
Qwen3 TTS
Qwen3 TTS
Alibaba's multilingual TTS with voice cloning, preset voices, and voice design from text.
1.7B 2025
— まだ投票はありません
medium Standard

詳細なベンチマーク結果

公式のTTS.aiベンチマークは自然性，正確性，速度の三つの次元で得点を得た。

KokoroKokoro

Free
自然性 4.8/5
精度 4.7/5
スピード 4.9/5
総合 4.8/5

CosyVoice 2CosyVoice 2

Standard
自然性 4.5/5
精度 4.4/5
スピード 3.8/5
総合 4.26/5

ChatterboxChatterbox

Premium
自然性 4.7/5
精度 4.5/5
スピード 3.4/5
総合 4.25/5

StyleTTS 2StyleTTS 2

Premium
自然性 4.5/5
精度 4.3/5
スピード 3.8/5
総合 4.23/5

PiperPiper

Free
自然性 3.5/5
精度 4.2/5
スピード 4.95/5
総合 4.15/5

MeloTTSMeloTTS

Free
自然性 3.8/5
精度 4.1/5
スピード 4.6/5
総合 4.13/5

Dia TTSDia TTS

Standard
自然性 4.6/5
精度 4.3/5
スピード 3.2/5
総合 4.09/5

VITSVITS

Free
自然性 3.4/5
精度 4.0/5
スピード 4.8/5
総合 4.0/5

OrpheusOrpheus

Standard
自然性 4.3/5
精度 4.1/5
スピード 3.5/5
総合 4.0/5

OpenVoiceOpenVoice

Premium
自然性 4.0/5
精度 4.1/5
スピード 3.9/5
総合 4.0/5

IndexTTS-2IndexTTS-2

Standard
自然性 4.3/5
精度 4.1/5
スピード 3.2/5
総合 3.91/5

Spark TTSSpark TTS

Standard
自然性 4.2/5
精度 4.0/5
スピード 3.4/5
総合 3.9/5

Parler TTSParler TTS

Standard
自然性 4.1/5
精度 3.9/5
スピード 3.4/5
総合 3.83/5

Tortoise TTSTortoise TTS

Premium
自然性 4.6/5
精度 4.4/5
スピード 1.8/5
総合 3.7/5

BarkBark

Standard
自然性 4.2/5
精度 3.8/5
スピード 2.5/5
総合 3.57/5

ベンチマーク方法論

テスト設定

  • ハードウェア: 4x NVIDIA Tesla P40（各24GB VRAM）、合計96GB
  • テストテキスト: 異なる言語パターン（語り、対話、技術的、感情的、多言語）をカバーする5つの標準化されたパターン
  • 評価: 自動測定（ＭＯＳ推定，ＷＥＲ，ＲＴＦ）と人間の聴取テストを組み合わせた。
  • 実行: 各モデルを１回に１０回テストし，得点を平均した。

得点基準

  • 自然さ(40%)： Prosody, intonation, rhythm, emotion — how human does it sound?
  • 精度 (30%): Pronunciation correctness, word error rate, intelligibility
  • 速度 (30%): リアルタイム因子 (オーディオ秒/生成秒)。高いほど速い。
  • 総合: 加重平均: 0.4 x 自然度 + 0.3 x 正確度 + 0.3 x スピード

ノート: Benchmarks reflect performance on our specific hardware and test texts. Real-world quality may vary based on input text, language, and voice selection. Community ratings provide a complementary signal based on diverse real usage.

よくある質問

The TTS Arena is a leaderboard that ranks AI text-to-speech models based on official benchmark tests and community ratings. Compare models side-by-side, listen to samples, and vote for the ones that sound best to you.

We run standardized tests on each model using the same text passages, hardware, and evaluation criteria. Scores cover naturalness (how human it sounds), accuracy (pronunciation and intelligibility), and speed (generation time). All tests use our GPU server with NVIDIA Tesla P40 GPUs.

Yes! Click the stars next to any model to rate it from 1 to 5. You need to be signed in to vote. Your rating contributes to the community average shown on the leaderboard. You can change your rating at any time.

Type any text, select two models, and click Compare. Both models generate speech from the same text simultaneously. Listen to both and vote for which sounds better. This blind comparison helps identify the best model for your specific needs.

Naturalness measures how human-like the speech sounds (prosody, intonation, rhythm). Accuracy measures pronunciation correctness and intelligibility. Speed measures how fast the model generates audio relative to real-time. Overall is a weighted average of all metrics.

Models without benchmark scores are either newly added and awaiting testing, or require special setup (like gated access tokens) that is pending. Community ratings are still available for these models.

Official benchmarks are updated when models receive significant updates or when new models are added. Community ratings update in real-time as users vote. The leaderboard data is cached for 5 minutes for performance.

Free models (Kokoro, Piper, VITS, MeloTTS) cost 0 credits. Standard models cost 2 credits per 1,000 characters. Premium models cost 4 credits per 1,000 characters and generally offer the highest quality or unique features like voice cloning.

For most use cases, Kokoro (free tier) offers excellent quality. For voice cloning, try Chatterbox or CosyVoice 2. For multilingual content, MeloTTS or CosyVoice 2. For expressive narration, Bark or Dia. Use the comparison tool to test with your specific text.

Yes, you can generate and compare audio from any two models without an account using free-tier models. Voting on models requires a free account. Premium model comparisons require credits.

We strive for objectivity by using standardized test texts, identical hardware, and consistent evaluation criteria across all models. Community ratings provide an additional independent signal. Our methodology is described in the Benchmark Methodology section below.

Models are ranked primarily by official benchmark overall score, then by community average rating as a tiebreaker. Models without benchmarks are ranked below those with benchmarks, ordered by community rating.
5.0/5 (1)

ファインド・ユア・パーフェクト・ボイス

Kokoro、Piper、VITS、MeloTTSでどのモデルでも無料で試してみてください。アカウントは必要ありません。

無料登録 価格を表示