KCharselect unicode block name

MIT, Apache 2.0 - ba a dakatar da shi ba, ba a dakatar da amfani da shi ba, ba a dakatar da biyan kuɗi ba. Yi amfani da su ta hanyar API ɗinmu mai zaman kansa, ko kuma ku shirya su a kan ginin ku na kanka tare da cikakken kulawa.

QSoftKeyManager QFontDatabase Apache KCharselect unicode block name GitHub

@ action

Free with Kokoro, Piper, VITS, MeloTTS
Za'a nuna sauti da ka samar a nan
@ action
QFileDialog
Yaushe kake son TTS.ai? Ka gaya wa abokanka!

@ item Spelling dictionary

Me yasa kayan aikin da aka bude suna da muhimmanci ga ayyukanka

All Open-Source Licensed

Duk wani nau'i a kan TTS.ai yana amfani da lasisin mai sauki mai sauki. Babu akwatunan baƙi masu mallaka, babu mai siyar da kuskure, babu kuɗin lasisin da ba a tsammani ba.

Apache

An ba da lasisi ga ma'aurata a karkashin MIT ko Apache 2.0, mafi yawan lasisi masu sauki. Yi amfani da su a cikin kasuwanci, canza su, raba su — babu iyaka.

KCharselect unicode block name

Ka saukar da kowanne nau'i kuma ka yi shi a kan kayan aikinka. Ka yi iko da duk bayananka, da kuma tsarinka. Babu bukatar dogaro da giciye.

QSoftKeyManager

An inganta siffofin ga NVIDIA GPUs tare da goyon bayan CUDA. Piper yana tafiya a kan CPU kawai. Mafi yawan siffofin suna buƙatar 2-8GB VRAM don samun sakamako mai kyau.

QDialogButtonBox

Jami'o'i masu sauki masu sauki suna kiyayewa da inganta waɗannan nau'ikan. Tallace-tallace suna maraba — gabatar da kurakurai, ingantawa, da sauti masu sauki a GitHub.

QPrintPreviewDialog

Dukan nau'ikan suna yarda da amfanin kasuwanci a ƙarƙashin lasisinsu. Gina kayayyakin aiki, sayar da sabis, da kuma ƙirƙirar abun ciki na kasuwanci ba tare da biyan kuɗi ko amfani da kuɗi ba.

QPrintPreviewDialog

Duk wani nau'i, lasisi, da abin da yake yi mafi kyau

KokoroKokoro

Free

Lightweight 82M parameter model delivering studio-quality speech with blazing-fast inference.

Fast 5/5

Mafi kyawun ga: Apache 2. 0 — mafi kyawun ingancin free model, 82M params, mai sauki ga self- host

QDialogButtonBox Kokoro

PiperPiper

Free

A fast, local neural text to speech system optimized for Raspberry Pi and embedded devices.

Fast 3/5

Mafi kyawun ga: MIT - CPU- kawai, cikakke ga na'urorin gefe da embedded self-hosting

QDialogButtonBox Piper

VITSVITS

Free

Conditional variational autoencoder with adversarial learning for end-to-end text-to-speech.

Fast 3/5

Mafi kyawun ga: MIT — tsarin ginin tushe wanda ake amfani dashi da dama daga cikin ma'aunin ƙasa

QDialogButtonBox VITS

BarkBark

Standard

Transformer-based text-to-audio model that generates realistic speech, music, and sound effects.

Slow 4/5

Mafi kyawun ga: MIT — iyakoki na musamman na samar da sauti a baya ga TTS na ƙa'ida

QDialogButtonBox Bark

Tortoise TTSTortoise TTS

Premium

Multi-voice text-to-speech focused on quality with autoregressive architecture.

Slow 5/5 QShortcut

Mafi kyawun ga: Apache 2. 0 - mafi kyawun inganci, aikace-aikacen alaƙa da aka koya sosai

QDialogButtonBox Tortoise TTS

OpenVoiceOpenVoice

Premium

Instant voice cloning with granular control over style, emotion, and accent.

Medium 4/5 QShortcut

Mafi kyawun ga: MIT — mai ƙãga halittar sauti mai maɓallin maɓalli mai maɓalli

QDialogButtonBox OpenVoice

Yadda za a Yi Amfani da Open Source TTS

Yi amfani da API ɗinmu ko kuma tafiyar da nau'ikan ka kaɗai

1

QPrintPreviewDialog

Browse our catalog of 20+ open-source TTS models. Each model page shows the license, architecture, capabilities, and self-hosting requirements.

2

QDialogButtonBox

Yi gwajin kowane nau'i kai tsaye akan TTS.ai ba tare da shigar da komai ba. Masu ba da sabis na GPU namu suna kula da sarrafawa don haka zaku iya tantance ingancin kafin ku yi alkawarin yin ajiyar kanka.

3

@ action

Clone model repos daga GitHub da kuma tafiyar da cikin gida, ko amfani da mu hosted API ga samarwa. Self-hosting ba da cikakken iko; mu API ba da gudanar da infrastructural.

4

Shirin AyukaName

Integration TTS cikin samfurinka ta amfani da self-hosted models ko mu REST API. Dukkanin samfuran suna amfani da kasuwanci ba tare da biyan kuɗi ko biyan kuɗi ba.

QFileDialog

Dukkan sifofin kan TTS.ai suna amfani da lasisin ma'adinai mai sauki

@ action QFileDialog QFontDatabase @ action QDialogButtonBox QShortcut
Kokoro Apache 2.0 @ action
Piper MIT QShortcut
VITS MIT QShortcut
MeloTTS MIT QShortcut
Chatterbox MIT QShortcut
Tortoise TTS Apache 2.0 @ action
StyleTTS 2 MIT QShortcut
OpenVoice MIT QShortcut
Sesame CSM Apache 2.0 @ action
Orpheus Llama 3.2 "Built with Llama"

QDialogButtonBox

Run models yourself or let us handle the infrastructure

QDialogButtonBox

Duk wani nau'i a kan TTS.ai yana samuwa a matsayin wani shirin asali mai budewa a kan GitHub ko Hugging Face. Sauke nauyin, shigar da dangantaka, da kuma gudanar da inference a kan GPUs ɗinka. Kana da cikakken iko akan latency, sirri, da kuma girma.

  • Full data privacy — audio never leaves your server
  • Ba'a da kudin kowace tambaya bayan daidaitar farko
  • @ action
  • Buƙaci kayan aiki na GPU (an shawarci NVIDIA)
  • Kuna kula da sabuntawa, girma, da dangantaka

Yi amfani da TTS.ai Hosted API

Get instant damar zuwa dukan 20+ models ta hanyar daya REST API. Mu kula da GPU bayarwa, model sabuntawa, kula da jerin gwano, da kuma girma. daya API maɓalli yana ba ka damar samun damar zuwa kowane model - ba bukatar gudanar da daban-daban aikace-aikace.

  • Ba'a bukata kayan haɗi na GPU ba
  • Duk 20+ siffofin ta hanyar API daya
  • QDialogButtonBox
  • 99.9% uptime tare da tsarin aiki mai yawa
  • Bã ku biyan kõme fãce abin da kuke aikatãwa.

QDialogButtonBox

Yi amfani da API ɗinmu mai ɗaurawa, ko shigar da Kokoro a cikin mintina

Zaɓi 1: TTS.ai Hosted API QPrintPreviewDialog
import requests

response = requests.post("https://api.tts.ai/v1/tts", json={
    "text": "Open source TTS with a simple API.",
    "model": "kokoro",
    "voice": "af_heart",
    "format": "wav"
}, headers={"Authorization": "Bearer YOUR_API_KEY"})

with open("output.wav", "wb") as f:
    f.write(response.content)
Zaɓuɓɓuka 2: Mai Gudanar da Kanka da pip QDialogButtonBox
# Install Kokoro locally
pip install kokoro

# Generate speech on your own GPU
import kokoro

pipeline = kokoro.KPipeline(lang_code="a")
generator = pipeline("Hello from your own server!", voice="af_heart")
for i, (gs, ps, audio) in enumerate(generator):
    kokoro.save(audio, f"output_{i}.wav")

Open Source, Affordable Pricing

API ɗinmu mai ɗaurawa yana sa TTS mai ma'ana mai budewa ya zama mai sauƙin isarwa ba tare da kula da GPU ba.

QPrintPreviewDialog

$0

@ action

  • 4 open-source models free
  • Ba a shiga don amfanin farko ba
  • An yarda da amfanin kasuwanci

@ action

$9

500,000 characters/month

  • All 20+ open-source models
  • @ action
  • API

QShortcut

$29

2,000,000 characters/month

  • GPU mai kula da fifiko
  • All premium models
  • QShortcut
View Full Pricing

Tambayar da ake yi da yawa

Tambayoyi masu yawa game da rubutun ma'anar maɓalli mai sauki zuwa magana

Na'am. Duk wani nau'i a kan TTS.ai yana amfani da lasisin ma'ana mai budewa — ko dai MIT ko Apache 2.0. Muna dauke da nau'i-nau'i da lasisin hanawa (kamar CPML na Coqui ko CC-BY-NC ba na kasuwanci ba). Za ka iya tabbatar da lasisin kowane nau'i a kan ajiyar GitHub.

Su biyun sun kasance lasisi masu sauki na maɓuɓɓugar maɓalli waɗanda ke ba da damar amfanin kasuwanci, canjawa, da kuma sake rabawa. Apache 2. 0 yana ƙara kyautar lasisi na bayyananne kuma yana buƙatar bayyana sauye- sauyen idan kun canza shiri. MIT yana da sauƙi tare da buƙatu kaɗan. Su biyun suna da kyau ga kasuwanci.

Na'am. Duk wani nau'i na iya zama mai zaman kansa. Clone da wurin ajiyar nau'i daga GitHub, shigar da dangantaka, sauke nau'in nau'i, da kuma tafiyar da inference. Muna bayar da takardun shaida ga bukatun kowane nau'i na mai zaman kansa ciki har da GPU, RAM, da kuma sigar Python.

Ka'idoji suna canzawa dangane da nau'in. Piper ba ya buƙatar GPU (CPU kawai). Kokoro da MeloTTS suna buƙatar 1-2GB VRAM. Mafi yawan nau'ikan standard suna buƙatar 4GB VRAM. Tortoise da Sesame CSM suna buƙatar 8GB. An NVIDIA RTX 3060 (12GB) na iya tafiyar da mafi yawan nau'ikan cikin sauƙi.

Yes. Open-source licenses allow modification including fine-tuning. Models like GPT-SoVITS and Bark provide fine-tuning scripts. You can train models on your own voice data to create custom voices or improve performance for specific languages.

Top open-source models (Kokoro, StyleTTS 2, Chatterbox) yanzu sun haɗu ko suka wuce sabis na kasuwanci kamar ElevenLabs da Google TTS a cikin ma'aunin inganci. Babban fa'idar sabis na kasuwanci shine tsarin da aka tsara da goyon baya, ba ingancin sauti ba.

Mun riga mun cire su. XTTS/XTTS-v2 (Coqui's CPML — ba-komercial), F5-TTS (CC-BY-NC — ba-komercial), da Higgs-v2 (Boson License — restrictive) an cire su duka. Duk wani nau'i akan TTS.ai an tabbatar da shi yana da aminci ga amfanin kasuwanci.

Na'am. Mafi yawan ma'aurata suna karɓar gudummawar jama'a ta hanyar GitHub. Za ka iya aika rahoton kurakurai, rikodin sauti ga yarukan da suka fi dacewa, inganta kodin, da takardun shaida. Ka duba repository na GitHub na kowace ma'aurata don sharuɗɗan gudummawa da matsalolin da ke aiki.

Load models on demand and unload when idle to share GPU memory. Our GPU server runs 20+ models on 4x Tesla P40 (96GB total VRAM) using dynamic loading. For self-hosting, a single 24GB GPU can serve 3-5 models concurrently.

Wasu nau'ikan suna bayar da hotunan Docker na hukuma ko fayilolin Docker. Don gudanar da nau'ikan da yawa, zaku iya gina tsarin Docker na musamman tare da NVIDIA Container Toolkit don samun damar GPU. Tsarin tsarin mai ba da sabis na API namu na iya aiki kamar wata ma'ana.

Mafi yawan nau'ikan suna buƙatar Python 3.10-3.12. Coqui TTS (VITS) musamman yana buƙatar Python 3.11. Muna shawartar Python 3.12 ga nau'ikan da yawa. Bincika kowane nau'in requirements.txt don daidaitaccen daidaitaccen sigar.

Na'am. MIT da Apache 2.0 lasisi bayyanannu yarda amfanin kasuwanci. Za ka iya gina SaaS kayayyakin, mobile apps, wasanni, da sabis amfani da wadannan models da babu lasisi kudin, royalties, ko bukatar shaida (ko da yake shaida ne godiya).
5.0/5 (1)

@ info

@ action

20+ open-source models, duk da lasisin kasuwanci. Yi amfani da API ko mai zaman kansa - zaɓin shine na ku.