OpenModelsHQ

NVIDIA-Nemotron-3.5-Transcribe-30B-A4B

NVIDIASpeechMoECustom license

NVIDIA-Nemotron-3.5-Transcribe-30B-A4B is an open-weight speech to text model from NVIDIA with 33.7B parameters (mixture of experts). At 4-bit it needs about 24 GB of memory, so the smallest current device that fits it is the Radeon AI PRO R9700.

Overview

Estimated memory at 4-bit
24GB

Estimated memory for the weights at 4-bit (plus runtime overhead). Image, video and speech models also need working memory that depends on resolution or clip length.

Total parameters
33.7B
Mixture of experts, all weights counted
Type
Speech to text
Released 7 Oct 2026
Cheapest API, blended / 1M
No hosted API
Weights only

Key information

Lab
NVIDIA
Released
7 Oct 2026
Parameters
33.7B · MoE
Type
Speech to text
HF downloads, 30 days
78,023
View all

Will it run on your hardware?

How we estimate
Memory needed vs devices
NVIDIA-Nemotron-3.5-Transcribe-30B-A4B needs about 24 GB at 4-bit, 42.4 GB at 8-bit and 80.1 GB at 16-bit. At 4-bit it fits GeForce RTX 5090, DGX Spark, Mac Studio M5 Ultra (256GB). Dashed lines are usable memory on each device; a model fits when its estimate stays within 90% of it. Log scale. 8k context.
Memory needed vs devices: data
Estimate or deviceMemory
Q424 GB
Q842.4 GB
BF1680.1 GB
RTX 5060 Ti 16 GB16 GB usable
RTX 5090 32 GB32 GB usable
DGX Spark 128 GB108.8 GB usable
Mac Studio M5 Ultra 256 GB192 GB usable

Single-device memory estimate. Pick a device to update the On your device column.

36 GB usable memory

FormatMemory neededOn your deviceExample hardware
4-bit (Q4)≈ 24 GBFits 67%Radeon AI PRO R970032 GB memory
8-bit (Q8)≈ 42.4 GBToo big 118%L40S48 GB memory
16-bit (BF16)≈ 80.1 GBToo big 222%RTX PRO 6000 Blackwell96 GB memory

Fits means the estimate stays within 90% of usable memory. Tight is above 90% and up to 100%. Too big is above 100%. These are memory checks, not speed or compatibility tests.

Mac allowances are 66% of total RAM up to 36 GB, and 75% above. AI boxes allow 85% because their memory is shared with the operating system.

Where to run it

On a Mac

No trusted MLX build found yet.

Check your Mac

With GGUF

No trusted GGUF build found yet.

Find compatible hardware

Through an API

No hosted API found.

More about this model

Frequently asked questions

How much memory does NVIDIA-Nemotron-3.5-Transcribe-30B-A4B need? About 24 GB at 4-bit, 42.4 GB at 8-bit and 80.1 GB at 16-bit (BF16), for the weights plus runtime overhead.

Can I run NVIDIA-Nemotron-3.5-Transcribe-30B-A4B on a Mac? The smallest current Mac that fits the 4-bit estimate of 24 GB is the Mac mini M5 Pro (48GB). We have not found a trusted MLX or GGUF build yet.

What does it need to run? The weights need about 24 GB at 4-bit and 80.1 GB at 16-bit. Image, video and speech models also need working memory that depends on resolution or clip length, so leave extra room. We have not found a trusted MLX or GGUF build yet.

What does it cost through an API? We did not find a hosted API for this model.

License in plain English

Custom license. This model uses a custom license rather than a standard one. Read it on the model card before using the model commercially. This summary is not legal advice.

Read the license on the model card

OMHQ RigBench

Not yet tested. We show no speed or quality score without a measurement.

Revision history and sources

Last updated . Sources:Hugging Face model card.