Free AI Text to SpeechQuery

82M ma parameters Mofulumira kwambiri Maganizo owoneka bwino Chilankhulo chosiyanasiyana Streaming thandizo

Lightweight 82M parameter model imabweretsa mawu a studio-quality ndi chidziwitso chofulumira.

Fast · 1.5GB VRAM Yambitsani

Piper

CPU-friendly Opanda zingwe zogwirizana 100 + mawu 30 + zilankhulo Kuthandizira SSML

M'malo mwake, njira yosinthira mawu ndi mawu yosinthidwa yosinthidwa kwa Raspberry Pi ndi zida zophatikizidwa.

Fast · 0 (CPU only) VRAM Yambitsani

VITS

End-to-end synthesization Malemba a m'Baibulo Fast kumvetsa Okamba ambiri

Kusintha kwa autoencoder ndi kuphunzira kwa adversarial kwa end-to-end text-to-speech.

Fast · 1GB VRAM Yambitsani

MeloTTS

CPU-optimized Mabungwe Malemba ambiri Kutulutsa-kuthamanga Low latency

High-quality multilingual text-to-speech yomwe imagwira ntchito pa CPU ndi latency yochepa.

Fast · 0.5GB (GPU optional) VRAM Yambitsani

Bark

Zochitika za mawu Kucheka / kupweteka Music chitukuko 100 + okamba Chilankhulo chosiyanasiyana

Model ya text-to-audio yokhala ndi transformer yomwe imapanga mawu, nyimbo, ndi zotsatira za mawu zowoneka bwino.

Slow · 5GB VRAM Yambitsani

Bark Small

Lightweight Mofulumira kuposa Full Bark Chilankhulo cha Emotional Chilankhulo chosiyanasiyana

Lighter mtundu wa Bark ndi mofulumira kumvetsa ndi pansi kugwiritsa ntchito kumbukirani.

Medium · 2GB VRAM Yambitsani

CosyVoice 2

Kutumiza Zero-shot cloning Cross-lingual Kuwongolera maganizo Human-parity

Alibaba's scalable streaming TTS ndi khalidwe la munthu-parity ndi kupitilira-zero latency.

Dia TTS

Wolankhula ambiri Kukhazikitsa kwa dialog Kusinthasintha kwachilengedwe Kusonyeza maganizo 1.6B Parameters

Multi-wokamba nkhani dialogue chiyambi cha mtundu womwe amaumba zokambirana zachilengedwe pakati pa okamba nkhani.

Parler TTS

Kufotokozera kwa mawu Chilankhulo chachilengedwe kuwongolera Kulenga mawu osinthika Sizikufunikira mawu osankhidwa kale

Kufotokozera mawu mukufuna mu chilankhulo chachilengedwe ndi Parler amapanga mawu ogwirizana.

GLM-TTS

Lowest kulephera mtengo Kusintha kwa mawu Flow kugwirizana Malemba a m'Baibulo

Achire pansi kwambiri chizindikiro cha vuto kuchuluka pakati open-source TTS mafano.

IndexTTS-2

Kuwongolera maganizo Zero-shot Ma vector a maganizo Chilankhulo cha Expressive Fine-grained kuwongolera

Zero-shot TTS ndi kuwongolera kwa maganizo olimba komanso kutanthauzira kwakukulu.

Spark TTS

Kusintha kwa mawu Kuwongolera maganizo Kuwongolera kwa mtundu Kuchokera pa lamulo 5-second cloning

Kulankhula kloning TTS ndi controlable chisoni ndi kulankhula mtundu mwa kufunsa.

GPT-SoVITS

5-second cloning Kuimba mawu Kuphunzira kwa ma shots ochepa High fidelity Cross-lingual

Few-shot mawu kloning TTS kuti amachitanso chilichonse mawu kuchokera 5 masekondi a audio.

Slow · 6GB VRAM Yambitsani

Orpheus

Human-level emotion 100K maola ophunzitsa Mphatso yachilengedwe Chilankhulo cha Expressive

Model ya TTS yokhudzana ndi munthu yophunzitsa 100K maola a data ya mawu.

Chatterbox

Kusintha kwa zero-shot Kuwongolera maganizo High fidelity Kusintha kwa mtundu Kugwiritsa ntchito chitsanzo chimodzi

State-of-the-art zero-shot kujambula mawu ndi kuwongolera maganizo kuchokera ku Resemble AI.

Tortoise TTS

Mtengo wapatali Maganizo ambiri DALL-E chikhalidwe Kusintha kwa mawu Kusintha kwachidule

Multi-wolankhula malemba-ku-kulankhula kuganizira za katundu ndi autoregressive ukadaulo.

Slow · 8GB VRAM Yambitsani

StyleTTS 2

Mtundu wa munthu Kufalitsa kwa mtundu Maphunziro otsutsana Zosintha zachilengedwe High fidelity

Kusintha kwa mawu kukhala mawu pamalingaliro a munthu pogwiritsa ntchito kufalitsa kwa mtundu ndi kuphunzitsa motsutsana.

OpenVoice

Chipangizo 3: Kusintha kwa mawu Kuwongolera maganizo Kuwongolera kwa Accent Chilankhulo chosiyanasiyana

Instant mawu kloning ndi granular kuwongolera pa mtundu, chisoni, ndi accent.

Qwen3 TTS

Kusintha kwa mawu 9 preset mawu Kujambula mawu kuchokera pamutu Kuwongolera maganizo 10 azinenero

TTS ya Alibaba ndi mawu osiyanasiyana, mawu osankhidwa, ndi mawu opangidwa kuchokera ku malemba.

Medium · 7GB VRAM Yambitsani

Sesame CSM

Chilankhulo Timing yachilengedwe Kusinthasintha Backchannel 1B Parameters

Kulankhulana kwa mawu kumabweretsa uthenga woyenera ndi nthawi yoyenera ndi maganizo.

Slow · 8GB VRAM Yambitsani

Chatterbox Turbo

Sub-200ms latency Masamba 6x nthawi yachidule Kusintha kwa mawu Kulemba

Faster Chatterbox ndi sub-200ms latency ndi paralinguistic tags kwa laughs, kupweteka, ndi zina zambiri.

Fast · 2GB VRAM Yambitsani

Dia 2

Kutumiza kwa mkati Wolankhula ambiri Low latency Masamba a Paralinguistic Kufika 2 min kutulutsa

Kutumiza-pyamba conversational TTS ndi multi-wokamba nkhani uthenga ndi paralinguistic cues.

VoxCPM

44.1kHz audio Tokenizer-free Cross-lingual cloning Kutengera mfundo LoRA fine-tuning

TTS yopanda Tokenizer yomwe imapanga 44.1kHz audio ndi kugwirizana kwa masamba omvetsetsa.

OuteTTS

CPU inference Kumvetsa msakatuli Kusintha kwa mawu Ma backends ambiri Ma profiles a wokamba

LLM-ogwirizana TTS kuti amayenda pa CPU, GPU, kapena browser kudzera llama.cpp ndi Transformers.js.

Fast · 2GB VRAM Yambitsani

TADA

Zero hallucinations 5x mofulumira kuposa LLM TTS Kusonyeza maganizo 700s audio context Kusintha kwa 2

Zero-hallucination TTS ndi malemba-acoustic dual kugwirizana, 5x mofulumira kuposa kuyerekezera LLM TTS.

Fast · 5GB VRAM Yambitsani

VibeVoice

Wolankhula ambiri Kufika 90 min Podcast chiyambi Kugwirizana kwa wokamba 200ms kusonkhana

Microsoft model for long-form multi-speaker content monga podcasts ndi audiobooks.

Pocket TTS

100M ma parameters CPU inference Kusintha kwa mawu Kugwiritsa ntchito Kusintha

Model ya 100M ya Kyutai ndi kulumikizana kwa mawu kuchokera ku satifiketi imodzi.

Fast · 1GB VRAM Yambitsani

Kitten TTS

CPU-only inference Under 80MB model size 8 built-in voices Speed control ONNX-based 24kHz output

Ultra-lightweight TTS under 80MB. Runs on CPU without GPU.

Fast · 0GB VRAM Yambitsani

CosyVoice3

Bi-streaming Emotion control Voice cloning Speed/volume control Instruction following

Next-generation multilingual TTS with bi-streaming, emotion control, and zero-shot voice cloning.

MOSS-TTS

Ultra-long generation 20 languages Voice cloning Duration control Pronunciation control Code-switching

Ultra-long 20-language TTS supporting up to 1 hour of continuous generation with phoneme-level control.

Medium · 16GB VRAM Yambitsani

MegaTTS3