Model comparison
VibeVoice-Realtime-0.5B vs Qwen3-TTS-12Hz-0.6B-Base
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| VibeVoice-Realtime-0.5B | Qwen3-TTS-12Hz-0.6B-Base | |
|---|---|---|
| Parameters | 1.0B | 915M |
| Architecture | vibevoice-tts | qwen3-tts |
| Layers | — | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 3 | 10 |
| Smallest quantization | 0.65 GiB | 0.50 GiB |
| Q4_K_M | — | 4.72 GiB |
| Licence | mit | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | VibeVoice-Realtime-0.5B | Qwen3-TTS-12Hz-0.6B-Base | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |