Model comparison

Qwen3-TTS-12Hz-0.6B-Base vs OmniVoice

At Q4_K_M, OmniVoice is the smaller download — 659,959,328 bytes against 5,073,265,248.

From the file· summed bytes, KV per layer

Side by side

Qwen3-TTS-12Hz-0.6B-BaseOmniVoice
Parameters915M613M
Architectureqwen3-ttsomnivoice-lm
Layers
Native context
Mixture of expertsnono
Quantizations published107
Smallest quantization0.50 GiB0.56 GiB
Q4_K_M4.72 GiB0.61 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3-TTS-12Hz-0.6B-BaseOmniVoiceRatio
4,096
8,192
16,384
32,768
65,536
131,072