@ action KCharselect unicode block name Phonon:: MMF:: EffectFactory QShortcut QShortcut @ item Spelling dictionary Marketplace QPrintPreviewDialog
Report Bug / Feature Request

TTS Arena - AI Voice Model Leaderboard

Ka kwatanta 20+ zane-zane na rubutu zuwa magana. Ma'aunin aikace-aikace, darajar jama'a, da kuma kwatantawa ta gefe-da-gefe.

Yi rijista
Ba mu da sauti na TTS a cikin harshenka har yanzu. Ka taimake mu da ƙara naka! QShortcut

QPrintPreviewDialog

Ka rubuta rubutu, ka zaɓi nau'i biyu, kuma ka kwatanta sakamakon. Nau'in nau'i-nau'i na free-tier ba su buƙaci asusun ba.

Free models aiki ba tare da wani asusu. QSql don ya bambanta da manyan ma'aurata.

QPrintPreviewDialog

# @ action KCharselect unicode block name KCharselect unicode block name QDialogButtonBox QSoftKeyManager Dakata
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 QSql
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 QDialogButtonBox
medium Standard
3
Chatterbox
Chatterbox
State-of-the-art zero-shot voice cloning with emotion control from Resemble AI.
300M 2025
4.25 /5 QDialogButtonBox
medium Premium
4
StyleTTS 2
StyleTTS 2
Human-level text-to-speech through style diffusion and adversarial training.
100M 585h 2024
4.23 /5 QDialogButtonBox
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 QDialogButtonBox
fast Free
6
MeloTTS
MeloTTS
High-quality multilingual text-to-speech that runs on CPU with minimal latency.
25M 2024
4.13 /5 QDialogButtonBox
fast Free
7
Dia TTS
Dia TTS
Multi-speaker dialog generation model that creates natural conversations between speakers.
1.6B 2024
4.09 /5 QDialogButtonBox
medium Standard
8
VITS
VITS
Conditional variational autoencoder with adversarial learning for end-to-end text-to-speech.
25M 585h 2021
4.0 /5 QDialogButtonBox
fast Free
9
Orpheus
Orpheus
Human-level emotional TTS model trained on 100K hours of speech data.
3B 100000h 2025
4.0 /5 QDialogButtonBox
medium Standard
10
OpenVoice
OpenVoice
Instant voice cloning with granular control over style, emotion, and accent.
300M 2024
4.0 /5 QDialogButtonBox
medium Premium
11
IndexTTS-2
IndexTTS-2
Zero-shot TTS with fine-grained emotion control and high expressiveness.
300M 2025
3.91 /5 QDialogButtonBox
medium Standard
12
Spark TTS
Spark TTS
Voice cloning TTS with controllable emotion and speaking style via prompts.
500M 2025
3.9 /5 QDialogButtonBox
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 QDialogButtonBox
medium Standard
14
Tortoise TTS
Tortoise TTS
Multi-voice text-to-speech focused on quality with autoregressive architecture.
400M 50000h 2022
3.7 /5 QDialogButtonBox
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 QDialogButtonBox
slow Standard
16
Bark Small
Bark Small
Lighter version of Bark with faster inference and lower memory usage.
150M 100000h 2023
— QDialogButtonBox
medium Standard
17
GLM-TTS
GLM-TTS
Achieves the lowest character error rate among open-source TTS models.
300M 2025
— QDialogButtonBox
medium Standard
18
GPT-SoVITS
GPT-SoVITS
Few-shot voice cloning TTS that replicates any voice from just 5 seconds of audio.
200M 2024
— QDialogButtonBox
slow Standard
19
Qwen3 TTS
Qwen3 TTS
Alibaba's multilingual TTS with voice cloning, preset voices, and voice design from text.
1.7B 2025
— QDialogButtonBox
medium Standard
20
Sesame CSM
Sesame CSM
Conversational speech model generating natural dialogue with appropriate timing and emotion.
1B 2025
— QDialogButtonBox
slow Premium
21
Chatterbox Turbo
Chatterbox Turbo
Faster Chatterbox with sub-200ms latency and paralinguistic tags for laughs, coughs, and more.
350M 2025
— QDialogButtonBox
fast Standard
22
Dia 2
Dia 2
Streaming-first conversational TTS with multi-speaker dialogue and paralinguistic cues.
2B 2025
— QDialogButtonBox
fast Standard
23
VoxCPM
VoxCPM
Tokenizer-free TTS producing 44.1kHz audio with context-aware paragraph consistency.
500M 1800000h 2025
— QDialogButtonBox
fast Standard
24
OuteTTS
OuteTTS
LLM-based TTS that runs on CPU, GPU, or browser via llama.cpp and Transformers.js.
1B 5000h 2025
— QDialogButtonBox
fast Free
25
TADA
TADA
Zero-hallucination TTS with text-acoustic dual alignment, 5x faster than comparable LLM TTS.
1B 2026
— QDialogButtonBox
fast Standard
26
VibeVoice
VibeVoice
Microsoft's multi-speaker long-form TTS generating up to 90 minutes with 4 distinct speakers.
1.5B 100000h 2025
— QDialogButtonBox
fast Standard
27
Pocket TTS
Pocket TTS
Lightweight 100M parameter model by Kyutai with voice cloning from a single sample.
100M 50000h 2025
— QDialogButtonBox
fast Free
28
Kitten TTS
Kitten TTS
Ultra-lightweight TTS under 80MB. Runs on CPU without GPU.
80M 2025
— QDialogButtonBox
fast Free
29
CosyVoice3
CosyVoice3
Next-generation multilingual TTS with bi-streaming, emotion control, and zero-shot voice cloning.
500M 200000h 2025
— QDialogButtonBox
fast Standard
30
MOSS-TTS
MOSS-TTS
Ultra-long 20-language TTS supporting up to 1 hour of continuous generation with phoneme-level control.
8B 500000h 2026
— QDialogButtonBox
medium Premium
31
MegaTTS3
MegaTTS3
ByteDance's sparse alignment TTS with adjustable intelligibility vs. speaker similarity.
1B 100000h 2025
— QDialogButtonBox
slow Premium

QPrintPreviewDialog

Official TTS.ai benchmark scores across three dimensions: naturalness, accuracy, and speed.

KokoroKokoro

Free
QPrintPreviewDialog 4.8/5
QPrintPreviewDialog 4.7/5
QSoftKeyManager 4.9/5
QPrintPreviewDialog 4.8/5

CosyVoice 2CosyVoice 2

Standard
QPrintPreviewDialog 4.5/5
QPrintPreviewDialog 4.4/5
QSoftKeyManager 3.8/5
QPrintPreviewDialog 4.26/5

ChatterboxChatterbox

Premium
QPrintPreviewDialog 4.7/5
QPrintPreviewDialog 4.5/5
QSoftKeyManager 3.4/5
QPrintPreviewDialog 4.25/5

StyleTTS 2StyleTTS 2

Premium
QPrintPreviewDialog 4.5/5
QPrintPreviewDialog 4.3/5
QSoftKeyManager 3.8/5
QPrintPreviewDialog 4.23/5

PiperPiper

Free
QPrintPreviewDialog 3.5/5
QPrintPreviewDialog 4.2/5
QSoftKeyManager 4.95/5
QPrintPreviewDialog 4.15/5

MeloTTSMeloTTS

Free
QPrintPreviewDialog 3.8/5
QPrintPreviewDialog 4.1/5
QSoftKeyManager 4.6/5
QPrintPreviewDialog 4.13/5

Dia TTSDia TTS

Standard
QPrintPreviewDialog 4.6/5
QPrintPreviewDialog 4.3/5
QSoftKeyManager 3.2/5
QPrintPreviewDialog 4.09/5

VITSVITS

Free
QPrintPreviewDialog 3.4/5
QPrintPreviewDialog 4.0/5
QSoftKeyManager 4.8/5
QPrintPreviewDialog 4.0/5

OrpheusOrpheus

Standard
QPrintPreviewDialog 4.3/5
QPrintPreviewDialog 4.1/5
QSoftKeyManager 3.5/5
QPrintPreviewDialog 4.0/5

OpenVoiceOpenVoice

Premium
QPrintPreviewDialog 4.0/5
QPrintPreviewDialog 4.1/5
QSoftKeyManager 3.9/5
QPrintPreviewDialog 4.0/5

IndexTTS-2IndexTTS-2

Standard
QPrintPreviewDialog 4.3/5
QPrintPreviewDialog 4.1/5
QSoftKeyManager 3.2/5
QPrintPreviewDialog 3.91/5

Spark TTSSpark TTS

Standard
QPrintPreviewDialog 4.2/5
QPrintPreviewDialog 4.0/5
QSoftKeyManager 3.4/5
QPrintPreviewDialog 3.9/5

Parler TTSParler TTS

Standard
QPrintPreviewDialog 4.1/5
QPrintPreviewDialog 3.9/5
QSoftKeyManager 3.4/5
QPrintPreviewDialog 3.83/5

Tortoise TTSTortoise TTS

Premium
QPrintPreviewDialog 4.6/5
QPrintPreviewDialog 4.4/5
QSoftKeyManager 1.8/5
QPrintPreviewDialog 3.7/5

BarkBark

Standard
QPrintPreviewDialog 4.2/5
QPrintPreviewDialog 3.8/5
QSoftKeyManager 2.5/5
QPrintPreviewDialog 3.57/5

KCharselect unicode block name

@ action

  • Haɗin kai: 4x NVIDIA Tesla P40 (24GB VRAM kowace), 96GB gaba ɗaya
  • @ action 5 standardized passages covering different speech patterns (narrative, dialogue, technical, emotional, multilingual)
  • QSql @ info: status
  • Runs: An yi gwajin kowane nau'i sau 10 a kowace tafiya, an auna sakamakon

QShortcut

  • @ label: textbox Prosody, intonation, rhythm, emotions — how human does it sound?
  • QPrintPreviewDialog @ item Spelling dictionary
  • QPrintPreviewDialog Ma'aunin lokaci na gaskiya (dakika na sauti/dakika na samarwa). Mafi girma = mafi sauri.
  • QPrintPreviewDialog Median: 0.4 x Naturalness + 0.3 x Accuracy + 0.3 x Speed

Bayani: Benchmarks suna nuna aiki a kan kayan aikinmu na musamman da kuma rubutun gwaji. ingancin duniyar gaskiya na iya bambanta bisa ga rubutun shigarwa, harshe, da kuma zaɓin magana. ƙididdiga na al'umma suna ba da alamun haɗin gwiwa bisa ga amfanin gaskiya mai yawa.

Tambayar da ake yi da yawa

The TTS Arena ne a leaderboard da ranks AI rubutu-to- magana mizalai bisa ga ofishin benchmark gwaje-gwaje da jama'a ratings. Ka kwatanta mizalai gefe-da-gefe, saurari misali, da kuri'a ga waɗanda da sauti mafi kyau ga ka.

Muna gudanar da gwaje-gwaje masu inganci a kan kowane nau'i ta hanyar amfani da sassan rubutu masu kama da juna, kayan aiki, da kuma sharuɗɗan nazari. Jarrabawa suna rufe dabi'a (ta yaya mutum yake ji), daidaito (fassara da fahimta), da sauri (lokacin samarwa). Duk gwaje-gwajen suna amfani da mai ba da sabis ɗinmu na GPU tare da NVIDIA Tesla P40 GPUs.

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

@ action

@ label: textbox

@ info: status

An sabunta ma'aunin da aka yi amfani da su lokacin da ma'aunin ya samu sabuntawa masu muhimmanci ko kuma lokacin da aka ƙara ma'aunin da aka yi amfani da su. An sabunta darajar jama'a cikin lokaci mai zuwa yayin da masu amfani ke kada kuri'a. An adana bayanan leaderboard na minti 5 don aiki.

Free models (Kokoro, Piper, VITS, MeloTTS) cost 0 characters. Standard models use 2x characters (e.g., 1,000 characters of text costs 2,000 characters from your balance). Premium models use 4x characters and generally offer the highest quality or unique features like voice cloning.

Ga mafi yawan lokutan amfani, Kokoro (farashin mataki) yana ba da inganci mai kyau. Don ƙirƙirar sauti, yi kokarin Chatterbox ko CosyVoice 2. Ga abun ciki na harsuna da yawa, MeloTTS ko CosyVoice 2. Don magana mai bayyanawa, Bark ko Dia. Yi amfani da kayan aikin kwatantawa don gwadawa tare da rubutun ka na musamman.

Na'am, zaka iya samar da kuma kwatanta sauti daga kowane nau'i biyu ba tare da asusun ba ta amfani da nau'i-nau'i na free-tier. Zaɓe kan nau'i-nau'i yana buƙatar asusun kyauta. Premium model comparisons require characters.

Mun yi kokarin ga objectivity ta hanyar amfani da standardized gwajin rubutu, identical kayan aiki, da daidaitacce gwajin sharudda a kan dukan models. Community ratings bayar da wani tabbaci na musamman. Mu hanyar da aka bayyana a cikin Benchmark Methodology sashe kasa.

@ action: inmenu
5.0/5 (1)

@ info

QShortcut

Yi gwajin kowanne nau'i da Kokoro, Piper, VITS, ko MeloTTS. Babu bukata ga asusun.

Yi rijista QPrintPreviewDialog