Qwen3-TTS-12Hz-1.7B-CustomVoice alternatives

Qwen3-TTS-12Hz-1.7B-CustomVoice's smallest published quantization is 1.90 GiB. The models below do the same job with a different trade — less memory, comparable size, or a more permissive licence.

From the file· sizes from summed file bytes

Meaningfully smaller

under 70% of its smallest quantization
ModelParamsSmallestQuantsLicence
Ace-Step1.5160M0.04 GiB116other
Qwen3-TTS-12Hz-0.6B-Base915M0.50 GiB10apache-2.0
OmniVoice613M0.56 GiB7apache-2.0
orpheus-3b-0.1-ft3.8B0.91 GiB22apache-2.0
VieNeu-TTS-0.3B244M0.19 GiB2cc-by-nc-4.0
neutts-air748M0.43 GiB6apache-2.0
Fun-CosyVoice3-0.5B-25120.34 GiB3apache-2.0
chatterbox264M0.17 GiB37mit

Comparable in size

within ±40%, so a like-for-like swap
ModelParamsSmallestQuantsLicence
s2-proMoE4.6B2.40 GiB7other
VoxCPM22.3B1.57 GiB7apache-2.0
csm-1b1.6B1.34 GiB3cc
orpheus-3b-0.1-pretrained3.8B1.40 GiB45apache-2.0
Qwen3-TTS-12Hz-1.7B-VoiceDesign1.9B1.90 GiB2apache-2.0
VibeVoice-1.5B2.7B1.76 GiB20mit
VoxCPM-0.5B728M2.45 GiB4

More permissively licensed

licences that allow commercial use
ModelParamsSmallestQuantsLicence
Qwen3-TTS-12Hz-0.6B-Base915M0.50 GiB10apache-2.0
OmniVoice613M0.56 GiB7apache-2.0
orpheus-3b-0.1-ft3.8B0.91 GiB22apache-2.0
neutts-air748M0.43 GiB6apache-2.0
Fun-CosyVoice3-0.5B-25120.34 GiB3apache-2.0
VoxCPM22.3B1.57 GiB7apache-2.0
chatterbox264M0.17 GiB37mit
Kokoro-82M82M0.13 GiB4apache-2.0