What GPU do I need to run nvidia/nemotron-speech-streaming-en-0.6b?
618M parameters, published in F32. View on Hugging Face
Set up nemotron-sp…treaming-en-0.6bnemotron-speech-streaming-en-0.6b is published by nvidia on Hugging Face, with 223,288 downloads and 614 likes to date. It's a NemotronAsrStreamingForRNNT model built for automatic-speech-recognition, published natively in F32.
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) |
|---|---|---|---|---|---|
| FP32 | 2.3 GB | 2.8 GB | V100 | 1 | $0.088/hr |
| FP8 (quantized) | 0.6 GB | 0.7 GB | RTX 4070 Super | 1 | $0.121/hr |
| INT4 (quantized) | 0.3 GB | 0.3 GB | RTX 4070 Super | 1 | $0.121/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 nemotron-speech-streaming-en-0.6b at its published (F32) precision: 1× V100, at $0.088/hr per GPU ($0.088/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
nemotron-speech-streaming-en-0.6b: common questions
How much VRAM does nemotron-speech-streaming-en-0.6b need?
2.8 GB at FP32, 0.7 GB at FP8 (quantized), 0.3 GB at INT4 (quantized). All of those sit under 6 GB, the smallest capacity this page reasons about, so VRAM is not what limits where this model runs. The 2.3 GB of weights plus inference overhead is the whole requirement.
Can nemotron-speech-streaming-en-0.6b run in 16-bit instead of FP32?
Yes. Its published weights are FP32, 2.3 GB, or 2.8 GB once inference overhead is added. Loading the same weights in 16 bits halves that to 1.2 GB, or 1.4 GB with overhead. How much accuracy the cast costs is model-specific and is not measured here.
How many copies of nemotron-speech-streaming-en-0.6b fit on one V100?
5, by VRAM alone. That card carries 16.0 GB and one copy needs 2.8 GB at FP32, on a live rate of $0.088/hr for the whole card. Throughput is not modelled here, so 5 copies is not 5 times the requests served.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
Alternatives at this size
Other models for automatic-speech-recognition within about a third of nemotron-speech-streaming-en-0.6b's 618M parameters, from other model lines.
- whisper-large-v3-turbo (809M, F16)
- parakeet-tdt-0.6b-v3 (627M, F32)
- distil-large-v3 (756M, F16)
- Qwen3-ForcedAligner-0.6B (918M, BF16)
- Fun-ASR-Nano-2512-hf (830M, BF16)
More on nemotron-speech-streaming-en-0.6b
Fits on an 8 GB GPU at FP32: every model that fits in 8 GB.
Best speech-to-text models: how nemotron-speech-streaming-en-0.6b ranks against the rest.
Related reading: V100 pricing and specs, Whisper variants compared, and What AI inference is.
