ListenSpeedScoresCapabilities🗳 Vote ↗

Capabilities · 55 models

Every tracked model and what it can do. The capability toggles combine — check two and you see only models with both; the search box filters by name; click any column header to sort. Cross-lingual clone = clones a voice from a reference in one language and speaks a different one in that same voice (verified per-model from each model card/paper). The License column shows the exact license; the commercial toggle is a coarse "no NC / research-only clause" filter — check the license before relying on it.
ModelParamsReleasedTypeLanguagesCross-lingualExpressiveSample rateLicenseCPU
Chatterbox1.2BApr 2025clone——knob24kMIT✓
Chatterbox Turbo744MDec 2025clone——tags*24kMIT✓
Coqui XTTS-v2750MOct 2023clone✓ (17)✓—24kCPML (non-commercial)✓
CosyVoice 3 0.5B0.5BDec 2025clone✓✓desc24kApache 2.0—
Dia 1.6B-06261.6BJun 2025clone——tags44.1kApache 2.0—
dots.tts (soar)2BJun 2026clone✓ (24)✓—48kApache 2.0—
DramaBox3.3BApr 2026clone— (en)—desc48kLTX-2 Community (NC)—
Echo-TTS2.8BDec 2025clone——tags44.1kCC-BY-NC-SA 4.0—
F5-TTS v1330MOct 2024clone✓✓—24kCC-BY-NC✓
Fish Speech 1.5~500MNov 2024clone✓——44.1kCC-BY-NC-SA 4.0✓
Fish Speech S2-Pro4BMar 2026clone——tags44.1kResearch (non-commercial)—
Higgs Audio v3 TTS4BJun 2026clone✓ (100)✓tags24kResearch (NC)—
IndexTTS-21.5BJun 2025clone✓✓emo-ref + desc + knob24kApache 2.0✓
KittenTTS Nano 0.1<100MAug 2025preset—n/a—24kApache 2.0✓
Kokoro82MDec 2024preset✓n/a—24kApache 2.0✓
LFM2.5-Audio 1.5B1.5BDec 2025preset— (en)n/a—24kLFM Open v1.0✓
LongCat-AudioDiT 1B1.42BMar 2026clone✓ (zh+en)——24kMIT—
LongCat-AudioDiT 3.5B3.83BMar 2026clone✓ (zh+en)——24kMIT—
LuxTTS123MJan 2026preset—n/a—22.05kMIT✓
Magpie-TTS357MDec 2025preset✓ (9)n/aemotion voices*22.05kNVIDIA OML✓
Mars5-TTS1.2BJun 2024clone———24kAGPL-3.0✓
Maya13BOct 2025preset—n/atags + desc24kApache 2.0✓
MeloTTS~52MFeb 2024preset— (en)n/a—44.1kMIT✓
MetaVoice-1B1.2BFeb 2024clone———48kApache 2.0—
MioTTS 0.1B0.1BFeb 2026clone✓ (en+ja)——44.1kFalcon-LLM—
MioTTS 0.6B0.6BFeb 2026clone✓ (en+ja)——44.1kApache 2.0—
MiraTTS0.5BDec 2025clone——knob48kMIT—
Miso TTS 8B8.2BMay 2026clone— (en)——24kModified MIT—
MOSS-TTS v1.08BFeb 2026clone✓ (20)✓—24kApache 2.0—
MOSS-TTS v1.58BMay 2026clone✓ (31)✓tags (pause)24kApache 2.0—
MOSS-TTS-Nano100MApr 2026clone✓ (zh+en)——48kApache 2.0✓
NeuTTS Air748MSep 2025clone———24kApache 2.0✓
NeuTTS Nano229MDec 2025clone———24kApache 2.0✓
OmniVoice~1BMar 2026clone✓ (600+)✓tags*24kApache 2.0 code / CC-BY-NC weights✓
OpenVoice v2~100MApr 2024clone✓✓knob22.05kMIT✓
Orpheus-TTS 3B3.3BMar 2025preset— (en)n/atags24kApache 2.0—
OuteTTS 1.0 1B1BApr 2025both✓ (12)✓—44.1kCC-BY-NC-SA 4.0 + Llama 3.2✓
Parler-TTS Mini v1878MJun 2024preset—n/adesc*44.1kApache 2.0✓
Piper~25MBJan 2023preset✓n/a—22.05kGPL-3.0✓
Pocket-TTS100MJan 2026clone———24kApache 2.0✓
Qwen3-TTS 1.7B (CUDA-graph)1.7BJan 2026clone✓✓—24kMIT—
Qwen3-TTS 1.7B Base1.7BJan 2026clone✓✓—24kApache 2.0✓
Sesame CSM-1B1BMar 2025clone———24kApache 2.0✓
Soprano 1.1 80M80MJan 2026preset—n/a—32kApache 2.0✓
Step-Audio-EditX3BOct 2025clone——tags + desc24kApache 2.0—
StyleTTS 2~148MJun 2023clone——knob24kMIT✓
Supertonic 399MMay 2026preset✓ (31)n/atags24kMIT + OpenRAIL-M✓
VibeVoice 1.5B1.5BAug 2025clone———24kMIT✓
VibeVoice 7B7BSep 2025clone———24kMIT—
VibeVoice Realtime 0.5B0.5BDec 2025preset—n/a—24kMIT✓
VoxCPM2 2B2BApr 2026clone✓ (30)—desc48kApache 2.0✓
Voxtral 4B TTS4BNov 2025both✓——24kCC-BY-NC 4.0✓
WavTTS 0.67B0.67BMay 2026clone✓ (zh+en)——16kMIT code / CC-BY-NC 4.0 weights—
ZipVoice 123M123MJun 2025clone✓ (zh+en)——24kApache 2.0✓
Zonos v0.11.6BFeb 2025clone✓✓emo-ref + knob44.1kApache 2.0✓