Model comparison

chatterbox vs Kokoro-82M

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

chatterboxKokoro-82M
Parameters264M82M
Architecturechatterbox-s3genkokoro
Layers3
Native context
Mixture of expertsnono
Quantizations published374
Smallest quantization0.17 GiB0.13 GiB
Q4_K_M0.46 GiB
Licencemitapache-2.0

KV cache by context

the term that decides long-context viability
ContextchatterboxKokoro-82MRatio
4,096
8,192
16,384
32,768
65,536
131,072