TTS Arena — Muuqaalka Codka AI Leaderboard
La barbardhigo 20+ qoraal-in-dhagax-muuqaalka. Official benchmarks, ratings bulshada, iyo side-by-side la barbardhigo.
Ma lihin codadka TTS ee afkaaga weli. Na caawi inaad ku darto kuwaaga!
Iibso Codkaaga
Side-by-Side La barbardhigo
Tixraac qoraalka, dooro laba nooc, oo la barbardhigo natiijooyinka. Free-tier noocyada u baahan yihiin xisaab la'aan.
Moodooyinka bilaashka ah waxay shaqeeyaan iyada oo aan xisaab lahayn. Ku soo biir si ay u barbardhigaan qaababka premium.
Model Leaderboard
| # | Nooc | Shaqada | Bulshada | Qiimayntaada | Xawaaraha | Qiyaamaha |
|---|---|---|---|---|---|---|
| 1 |
Kokoro
Lightweight 82M parameter model delivering studio-quality speech with blazing-fast inference.
82M
1200h
2024
|
4.8 /5 |
5.0
/5
1 cod
|
fast | Free | |
| 2 |
CosyVoice 2
Alibaba's scalable streaming TTS with human-parity naturalness and near-zero latency.
300M
200000h
2024
|
4.26 /5 | Wax cod ah ma jiro | medium | Standard | |
| 3 |
Chatterbox
State-of-the-art zero-shot voice cloning with emotion control from Resemble AI.
300M
2025
|
4.25 /5 | Wax cod ah ma jiro | medium | Premium | |
| 4 |
StyleTTS 2
Human-level text-to-speech through style diffusion and adversarial training.
100M
585h
2024
|
4.23 /5 | Wax cod ah ma jiro | medium | Premium | |
| 5 |
Piper
A fast, local neural text to speech system optimized for Raspberry Pi and embedded devices.
15M
2023
|
4.15 /5 | Wax cod ah ma jiro | fast | Free | |
| 6 |
MeloTTS
High-quality multilingual text-to-speech that runs on CPU with minimal latency.
25M
2024
|
4.13 /5 | Wax cod ah ma jiro | fast | Free | |
| 7 |
Dia TTS
Multi-speaker dialog generation model that creates natural conversations between speakers.
1.6B
2024
|
4.09 /5 | Wax cod ah ma jiro | medium | Standard | |
| 8 |
VITS
Conditional variational autoencoder with adversarial learning for end-to-end text-to-speech.
25M
585h
2021
|
4.0 /5 | Wax cod ah ma jiro | fast | Free | |
| 9 |
Orpheus
Human-level emotional TTS model trained on 100K hours of speech data.
3B
100000h
2025
|
4.0 /5 | Wax cod ah ma jiro | medium | Standard | |
| 10 |
OpenVoice
Instant voice cloning with granular control over style, emotion, and accent.
300M
2024
|
4.0 /5 | Wax cod ah ma jiro | medium | Premium | |
| 11 |
IndexTTS-2
Zero-shot TTS with fine-grained emotion control and high expressiveness.
300M
2025
|
3.91 /5 | Wax cod ah ma jiro | medium | Standard | |
| 12 |
Spark TTS
Voice cloning TTS with controllable emotion and speaking style via prompts.
500M
2025
|
3.9 /5 | Wax cod ah ma jiro | medium | Standard | |
| 13 |
Parler TTS
Describe the voice you want in natural language and Parler generates matching speech.
880M
45000h
2024
|
3.83 /5 | Wax cod ah ma jiro | medium | Standard | |
| 14 |
Tortoise TTS
Multi-voice text-to-speech focused on quality with autoregressive architecture.
400M
50000h
2022
|
3.7 /5 | Wax cod ah ma jiro | slow | Premium | |
| 15 |
Bark
Transformer-based text-to-audio model that generates realistic speech, music, and sound effects.
350M
100000h
2023
|
3.57 /5 | Wax cod ah ma jiro | slow | Standard | |
| 16 |
Bark Small
Lighter version of Bark with faster inference and lower memory usage.
150M
100000h
2023
|
— | Wax cod ah ma jiro | medium | Standard | |
| 17 |
GPT-SoVITS
Few-shot voice cloning TTS that replicates any voice from just 5 seconds of audio.
200M
2024
|
— | Wax cod ah ma jiro | slow | Standard | |
| 18 |
Qwen3 TTS
Alibaba's multilingual TTS with preset voices and voice design from text.
1.7B
2025
|
— | Wax cod ah ma jiro | medium | Standard | |
| 19 |
VieNeu-TTS-v2
Vietnamese + English code-switching TTS with 7 preset voices and zero-shot voice cloning. CPU-only, no GPU required.
0.3B
10000h
2026
|
— | Wax cod ah ma jiro | fast | Standard | |
| 20 |
Sesame CSM
Conversational speech model generating natural dialogue with appropriate timing and emotion.
1B
2025
|
— | Wax cod ah ma jiro | slow | Premium | |
| 21 |
Chatterbox Turbo
Faster Chatterbox with sub-200ms latency and paralinguistic tags for laughs, coughs, and more.
350M
2025
|
— | Wax cod ah ma jiro | fast | Standard | |
| 22 |
VoxCPM
Tokenizer-free TTS producing 44.1kHz audio with context-aware paragraph consistency.
500M
1800000h
2025
|
— | Wax cod ah ma jiro | fast | Standard | |
| 23 |
Kani TTS 2
Ultra-lightweight 400M English TTS model running in just 3GB VRAM.
400M
10000h
2026
|
— | Wax cod ah ma jiro | fast | Free | |
| 24 |
OuteTTS
LLM-based TTS that runs on CPU, GPU, or browser via llama.cpp and Transformers.js.
1B
5000h
2025
|
— | Wax cod ah ma jiro | fast | Free | |
| 25 |
VibeVoice
Microsoft's multi-speaker long-form TTS generating up to 90 minutes with 4 distinct speakers.
1.5B
100000h
2025
|
— | Wax cod ah ma jiro | fast | Standard | |
| 26 |
Pocket TTS
Lightweight 100M parameter model by Kyutai with voice cloning from a single sample.
100M
50000h
2025
|
— | Wax cod ah ma jiro | fast | Free | |
| 27 |
Kitten TTS
Ultra-lightweight TTS under 80MB. Runs on CPU without GPU.
80M
2025
|
— | Wax cod ah ma jiro | fast | Free | |
| 28 |
CosyVoice3
Next-generation multilingual TTS with bi-streaming, emotion control, and zero-shot voice cloning.
500M
200000h
2025
|
— | Wax cod ah ma jiro | fast | Standard | |
| 29 |
NAMAA Saudi TTS
First open Saudi-Arabic TTS. Native Saudi dialect with Chatterbox-quality voice cloning.
300M
2026
|
— | Wax cod ah ma jiro | medium | Standard | |
| 30 |
Darwin TTS
Cross-modal Qwen3-TTS variant with FFN weights blended from the Qwen3-1.7B language model for sharper multilingual cloning.
2.1B
2026
|
— | Wax cod ah ma jiro | medium | Standard | |
| 31 |
MOSS-TTSD
Multi-speaker dialogue continuation model — generate podcast-style conversations with up to 5 speakers and 60 minutes of coherent audio.
7B
2026
|
— | Wax cod ah ma jiro | medium | Standard | |
| 32 |
Ming-Omni TTS
Compact 0.5B omni-modal speech model from inclusionAI with high-fidelity 44.1kHz output and zero-shot voice cloning.
500M
2026
|
— | Wax cod ah ma jiro | medium | Free | |
| 33 |
MOSS-TTS Nano
Tiny 100M MOSS-TTS variant — same architecture, 80x smaller, free-tier latency.
100M
500000h
2026
|
— | Wax cod ah ma jiro | fast | Free |
Dhammaan dhibcaha Benchmark
Sharciga TTS.ai benchmark dhibcaha saddex geesoodka ah: dabiiciga ah, saxnaanta, iyo xawaaraha.
Kokoro
Free
Caqliga
4.8/5
Xasaanad
4.7/5
Xawaaraha
4.9/5
Guud ahaan
4.8/5
CosyVoice 2
Standard
Caqliga
4.5/5
Xasaanad
4.4/5
Xawaaraha
3.8/5
Guud ahaan
4.26/5
Chatterbox
Premium
Caqliga
4.7/5
Xasaanad
4.5/5
Xawaaraha
3.4/5
Guud ahaan
4.25/5
StyleTTS 2
Premium
Caqliga
4.5/5
Xasaanad
4.3/5
Xawaaraha
3.8/5
Guud ahaan
4.23/5
Piper
Free
Caqliga
3.5/5
Xasaanad
4.2/5
Xawaaraha
4.95/5
Guud ahaan
4.15/5
MeloTTS
Free
Caqliga
3.8/5
Xasaanad
4.1/5
Xawaaraha
4.6/5
Guud ahaan
4.13/5
Dia TTS
Standard
Caqliga
4.6/5
Xasaanad
4.3/5
Xawaaraha
3.2/5
Guud ahaan
4.09/5
VITS
Free
Caqliga
3.4/5
Xasaanad
4.0/5
Xawaaraha
4.8/5
Guud ahaan
4.0/5
Orpheus
Standard
Caqliga
4.3/5
Xasaanad
4.1/5
Xawaaraha
3.5/5
Guud ahaan
4.0/5
OpenVoice
Premium
Caqliga
4.0/5
Xasaanad
4.1/5
Xawaaraha
3.9/5
Guud ahaan
4.0/5
IndexTTS-2
Standard
Caqliga
4.3/5
Xasaanad
4.1/5
Xawaaraha
3.2/5
Guud ahaan
3.91/5
Spark TTS
Standard
Caqliga
4.2/5
Xasaanad
4.0/5
Xawaaraha
3.4/5
Guud ahaan
3.9/5
Parler TTS
Standard
Caqliga
4.1/5
Xasaanad
3.9/5
Xawaaraha
3.4/5
Guud ahaan
3.83/5
Tortoise TTS
Premium
Caqliga
4.6/5
Xasaanad
4.4/5
Xawaaraha
1.8/5
Guud ahaan
3.7/5
Bark
Standard
Caqliga
4.2/5
Xasaanad
3.8/5
Xawaaraha
2.5/5
Guud ahaan
3.57/5
Habka la tixgelinayo
Istaraatiijiyad Imtixaan
- Qalabka: 4x NVIDIA Tesla P40 (24GB VRAM mid kasta), 96GB guud ahaan
- Imtixaan: 5 qaybaha caadiga ah oo ka kooban qaabab hadal kala duwan (warbixinta, wada hadalka, farsamada, dareenka, luqado badan)
- Qiimaha: Automatic tirakoobka (MOS tirakoobka, WER, RTF) oo la isku daray tijaabooyinka dhagaysiga aadanaha
- Socdaal: Nooca kasta oo la tijaabiyay 10 jeer oo isku mar ah, dhibcaha dhexdhexaad ah
Tilmaamaha Dhismaha
- dabiiciga ah (40%): Prosody, intonation, riiq, dareenka - sida aadanaha ma u egtahay?
- Xaqnimo (30%): Dhaqanka saxda ah, xadgudubka ereyga, fahanka
- Xawaaraha (30%): Real-wakhtiga faa'iido (audio seconds / generation seconds). Sare = ka dhaqso badan.
- Guud ahaan: Dhererka dhexdhexaadka ah: 0.4 x dabiiciga ah + 0.3 x saxnaanta + 0.3 x xawaaraha
Faahfaahin: Qiimaha dhabta ah ee adduunka ayaa kala duwanaan kara iyadoo ku xiran qoraalka soo dejinta, afka, iyo doorasho codka. Qiimaha bulshada waxay bixiyaan calaamad ku xiran isticmaalka dhabta ah ee kala duwan.
Su'aalaha badanaa la waydiiyo
Maxaa aan ku hagaajin karnaa? Jawaabtaada waxay naga caawisaa inaan xallino dhibaatooyinka.
Hesho Codkaaga Wanaagsan
Raac nooc kasta oo bilaash ah Kokoro, Piper, VITS, ama MeloTTS. Aan xisaab loo baahnayn.