How much VRAM does canary-180m-flash need?

Q4_K_M is exactly 139,223,744 bytes (0.13 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

nvidia · audio asr

canary-180m-flash

nvidia/canary-180m-flash

canary-180m-flash at Q4_K_M is exactly 139,223,744 bytes (0.13 GiB / 0.14 GB) — an effective 5.891 bits per weight, not the nominal 4.

From the file· summed from 1 file(s)

Parameters

189M

Architecture

canary

Context

—

native (config.json)

License

cc-by-4.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_K_M	0.13 GiB	139,223,744	5.891	789	handy-computer
Q5_K_M	0.15 GiB	158,704,320	6.715	789	handy-computer
Q6_K	0.16 GiB	176,291,520	7.459	789	handy-computer
Q8_0	0.20 GiB	218,447,552	9.242	789	handy-computer
F16	0.36 GiB	381,632,192	16.147	789	handy-computer
F32	0.70 GiB	756,498,112	32.007	—	handy-computer

Measured

published by a third party, attributed below

Metric	Value	What it means
rtf	2484.2
RTFx	2484.2	Higher is better — audio seconds processed per second of compute.
Word error rate	6.11%	Lower is better — the share of words transcribed incorrectly.
Word error rate	12.10%	Lower is better — the share of words transcribed incorrectly.
Word error rate	12.50%	Lower is better — the share of words transcribed incorrectly.
Word error rate	8.96%	Lower is better — the share of words transcribed incorrectly.
Word error rate	2.04%	Lower is better — the share of words transcribed incorrectly.
Word error rate	3.41%	Lower is better — the share of words transcribed incorrectly.
Word error rate	3.57%	Lower is better — the share of words transcribed incorrectly.
Word error rate	13.66%	Lower is better — the share of words transcribed incorrectly.
Word error rate	1.52%	Lower is better — the share of words transcribed incorrectly.
Word error rate	8.87%	Lower is better — the share of words transcribed incorrectly.
Word error rate	6.29%	Lower is better — the share of words transcribed incorrectly.

Benchmarked· by open-asr-leaderboard-english-short-latest

RTFx measured by the Open ASR Leaderboard on a single datacenter GPU at a large batch size. It ranks models against each other; it says nothing about throughput on consumer hardware. We reproduce these figures with attribution; they are not ours and we have not verified the runs. Source: open-asr-leaderboard-english-short-latest.

Compare with

same modality, comparable size

vs whisper-medium vs whisper-large-v3-turbo vs GigaAM-v3 vs whisper-small vs SenseVoiceSmall vs whisper-base All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 0.10 GiB. The real file is 0.13 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

Architecture unavailable — this repository is gated and no ungated mirror was found. Exact file sizes above are still authoritative; only the KV math needs the config.

Questions people ask

How much VRAM does canary-180m-flash need?: Q4_K_M is exactly 139,223,744 bytes (0.13 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of canary-180m-flash should I use?: Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.