What GPU do I need to run bosonai/higgs-audio-v3-8b-stt-v2?
8.9B parameters, published in BF16. View on Hugging Face
higgs-audio-v3-8b-stt-v2 is published by bosonai on Hugging Face, with 16,443 downloads and 14 likes to date. It's a HiggsAudio3Model model built for automatic-speech-recognition, published natively in BF16.
VRAM required & cheapest live GPU fit
Required VRAM = weight size at each precision, plus a fixed overhead for activations and allocator fragmentation. Speech models don't build the same growing KV-cache a text model does. Memory scales primarily with input audio length. Full formula and assumptions: methodology.
| Precision | Weight size | Required VRAM | Cheapest live fit | GPUs needed | Est. $/hr (full fit) |
|---|---|---|---|---|---|
| BF16 | 16.6 GB | 19.9 GB | RTX A5000 | 1 | $0.176/hr |
| FP8 (quantized) | 8.3 GB | 10.0 GB | RTX 5060 Ti | 1 | $0.180/hr |
| INT4 (quantized) | 4.1 GB | 5.0 GB | RTX 3070 | 1 | $0.088/hr |
A GPU is only matched to a row if its hardware supports that precision, and the primary recommendation is always a single-GPU fit when one exists.
INT4 requires a quantized checkpoint actually published for this model, check its Hugging Face page before relying on this row.
Cheapest way to run higgs-audio-v3-8b-stt-v2 at its published (BF16) precision: 1× RTX A5000, at $0.176/hr per GPU ($0.176/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
higgs-audio-v3-8b-stt-v2: common questions
Does higgs-audio-v3-8b-stt-v2 fit on a 24 GB GPU?
Yes. At BF16 it needs 19.9 GB of VRAM, so a 24 GB card holds it with 4.1 GB to spare. A 16 GB card is not enough for it at BF16.
What is the least VRAM higgs-audio-v3-8b-stt-v2 can run in?
5.0 GB, at INT4 (quantized), which fits a 6 GB card, against 19.9 GB at BF16. That row assumes an INT4 checkpoint has actually been published for this model, so check its Hugging Face page before planning around it.
Does quantizing higgs-audio-v3-8b-stt-v2 lower the GPU bill?
Yes. At BF16 the cheapest live fit is one RTX A5000 at $0.176/hr. At INT4 (quantized) it drops to one RTX 3070 at $0.088/hr, provided a quantized checkpoint exists for it.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More bosonai models
- higgs-tts-3-4b (4.7B, BF16)
Alternatives at this size
Other models for automatic-speech-recognition within about a third of higgs-audio-v3-8b-stt-v2's 8.9B parameters, from other model lines.
- VibeVoice-ASR (8.7B, BF16)
- granite-speech-3.2-8b (8.5B, BF16)
More on higgs-audio-v3-8b-stt-v2
Fits on a 24 GB GPU at BF16: every model that fits in 24 GB.
Fits on a 12 GB GPU at FP8: every model that fits in 12 GB.
Fits on an 8 GB GPU at INT4: every model that fits in 8 GB.
Best speech-to-text models: how higgs-audio-v3-8b-stt-v2 ranks against the rest.
Related reading: RTX A5000 pricing and specs, Whisper variants compared, and What AI inference is.
Your own GPU cloud accounts
One console for the GPU clouds you already rent from. Your keys stay on your machine.
$20 a month, flat