Model comparison

Qwen3-ASR-0.6B vs whisper-large-v3

At Q4_K_M, Qwen3-ASR-0.6B is the smaller download — 589,560,480 bytes against 997,303,008.

From the file· summed bytes, KV per layer

Side by side

Qwen3-ASR-0.6Bwhisper-large-v3
Parameters938M1.5B
Architectureqwen3_asrwhisper
Layers32
Native context
Mixture of expertsnono
Quantizations published1018
Smallest quantization0.55 GiB0.49 GiB
Q4_K_M0.55 GiB0.93 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3-ASR-0.6Bwhisper-large-v3Ratio
4,096
8,192
16,384
32,768
65,536
131,072