# NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 by NVIDIA: Open Model Source: https://savrn.com/models/nvidia-nemotron-3-nano-30b-a3b-nvfp4 Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 (18.2B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 36.5 GB | 43.8 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 18.2 GB | 21.9 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 9.1 GB | 10.9 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 23, 2026. [NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/nvidia-nemotron-3-nano-30b-a3b-nvfp4/gpus) ## SAVRN's Notes on NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 Start with the context: 262,144 tokens. NVIDIA built Nemotron 3 Nano as a single model for both reasoning and non-reasoning work: it writes a reasoning trace first, then the answer, and a flag in the chat template turns the reasoning on or off. This NVFP4 build is the quantized version of the BF16 release, with 128 routed experts. At 4-bit the weights are 9.1 GB and the run needs 10.9 GB, so the cheapest listed setup, one MI300X with 192 GB at $1.85 an hour, leaves most of the card for context. The license is listed as other with no summary on the page, so read NVIDIA's terms yourself before this goes near a product. Check the dates too: pre-training cutoff June 25, 2025, post-training November 28, 2025. The seven Nemotron training sets are named on the page, so you can see what went in. SAVRN Research, 2026-09-18 ## Model Card September 2025 \- December 2025 The post-training data has a cutoff date of November 28, 2025\. The pre-training data has a cutoff date of June 25, 2025\. Nemotron-Nano-3-30B-A3B-NVFP4 is a quantized version of Nemotron-Nano-3-30B-A3B and is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be configured through a flag in the chat template. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be… Excerpt from the card by NVIDIA, licensed other. ## Configuration Architecture NemotronHForCausalLM Context length (tokens) 262,144 Layers 52 Hidden size 2,688 Feed-forward size 1,856 Attention heads 32 Key/value heads 2 Head dimension 128 Vocabulary size 131,072 Routed experts 128 Experts active per token 6 RoPE base 10,000 Stored precision bfloat16 Model type nemotron_h ## Identity and Version Repository nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 Publisher NVIDIA Task Text generation Modality Text Library transformers Parameters 18.2B parameters Languages en, es, fr, de, ja, it Revision 6efb4a2a1c1fa277ce7b3df7a1416255011b1c99 First published 2025-12-20 Last updated 2026-08-24 ## Files and Weights 18 files, 19.4 GB in total. The weights are 5 files totalling 19.3 GB in safetensors. Weights5 files · 19.3 GB Configuration8 files · 2.6 MB Tokenizer2 files · 17.3 MB Documentation1 file · 73.7 KB Other1 file · 10.5 KB Repository1 file · 1.6 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model-00001-of-00005.safetensors | Weights | 4.0 GB | 2fdac76b3e49 | | model-00002-of-00005.safetensors | Weights | 4.0 GB | 559806ee0cb6 | | model-00003-of-00005.safetensors | Weights | 4.0 GB | d82084978870 | | model-00004-of-00005.safetensors | Weights | 4.0 GB | f5ccb7cfa787 | | model-00005-of-00005.safetensors | Weights | 3.3 GB | c9dd91428393 | | config.json | Configuration | 1.8 KB | — | | configuration_nemotron_h.py | Configuration | 12.9 KB | — | | generation_config.json | Configuration | 197 B | — | | hf_quant_config.json | Configuration | 3.0 KB | — | | model.safetensors.index.json | Configuration | 2.5 MB | — | | modeling_nemotron_h.py | Configuration | 83.8 KB | — | | nano_v3_reasoning_parser.py | Configuration | 798 B | — | | special_tokens_map.json | Configuration | 420 B | — | | README.md | Documentation | 73.7 KB | — | | chat_template.jinja | Other | 10.5 KB | — | | .gitattributes | Repository | 1.6 KB | — | | tokenizer.json | Tokenizer | 17.1 MB | c6021eb6847e | | tokenizer_config.json | Tokenizer | 188.0 KB | — | ## License and Download License other Access Open weights, no gate Download size 19.3 GB [Download from NVIDIA](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4) Released by NVIDIA through its official repository on Hugging Face. ## Built From - Derived from [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://savrn.com/models/nvidia-nemotron-3-nano-30b-a3b-bf16) - Described by [arXiv:2512.20848](https://savrn.com/papers/nemotron-3-nano-open-efficient-mixture-of-experts-hybrid-mamba-transformer-model-for-agent) - Described by [arXiv:2512.20856](https://savrn.com/papers/nvidia-nemotron-3-efficient-and-open-intelligence) - Described by arXiv:2601.20088 - Quantized from [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://savrn.com/models/nvidia-nemotron-3-nano-30b-a3b-bf16) - Trained on (disclosed) nvidia/Nemotron-3-Nano-RL-Training-Blend - Trained on (disclosed) nvidia/Nemotron-Agentic-v1 - Trained on (disclosed) nvidia/Nemotron-CC-Code-v1 - Trained on (disclosed) nvidia/Nemotron-CC-Math-v1 - Trained on (disclosed) nvidia/Nemotron-CC-v2 - Trained on (disclosed) nvidia/Nemotron-CC-v2.1 - Trained on (disclosed) nvidia/Nemotron-Competitive-Programming-v1 - Trained on (disclosed) nvidia/Nemotron-Instruction-Following-Chat-v1 - Trained on (disclosed) nvidia/Nemotron-Math-Proofs-v1 - Trained on (disclosed) nvidia/Nemotron-Math-v2 - Trained on (disclosed) nvidia/Nemotron-Pretraining-Code-v1 - Trained on (disclosed) nvidia/Nemotron-Pretraining-Code-v2 - Trained on (disclosed) nvidia/Nemotron-Pretraining-Dataset-sample - Trained on (disclosed) nvidia/Nemotron-Pretraining-SFT-v1 - Trained on (disclosed) nvidia/Nemotron-Pretraining-Specialized-v1 - Trained on (disclosed) nvidia/Nemotron-Science-v1 ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 19.3 GB | | 16-bit | 36.5 GB | | 8-bit | 18.2 GB | | 4-bit | 9.1 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Compare NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 - [NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 vs Qwen3.6-35B-A3B-NVFP4](https://savrn.com/models/compare/nvidia-nemotron-3-nano-30b-a3b-nvfp4-vs-qwen3-6-35b-a3b-nvfp4) ## Questions About NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 ### How much GPU memory does NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 need? About 43.8 GB at 16-bit and 10.9 GB at 4-bit: the weights (18.2B parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### What license is NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 released under? other, as its publisher declares it. Read the license text before commercial use. ### What is NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4's context length? 262,144 tokens, from the maximum position embeddings in its published configuration. ## Similar Models Model · Text generation ### [veriloop-coder-e2-nvfp4](https://savrn.com/models/veriloop-coder-e2-nvfp4) [Rodrigo Ramos Da Silveira](https://savrn.com/model-publishers/rodrigoramosrs) Apache-2.0 This repository contains an NVFP4 quantization of VeriLoop E2, an open 27B post-trained model built on Qwen3.8-27B for code, mathematics, and physics. Its core reasoning discipline is VeriLoop-Governed Recurrence (VGR): candidate states are recursively proposed, externally checked, and retained only when the protected evidence state improves without regression. Quantized by Rodrigo Ramos. Produced with NVIDIA Model Optimizer using the canonical NVFP4W4A4WEIGHTLOCALHESSIANCFG recipe: static per-block 4-bit weights (group size 16) + dynamic 4-bit activations, FP8 attention, local-Hessian calibration with MSE and fp8 scale sweep. The linearattn (Gated Delta Net) blocks and the… Open weights apache-2.0 18.3B parameters 262,144 tokens transformers [View model](https://savrn.com/models/veriloop-coder-e2-nvfp4) Model · Text generation ### [NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4](https://savrn.com/models/nvidia-nemotron-3-5-lightning-30b-a3b-nvfp4) [NVIDIA](https://savrn.com/model-publishers/nvidia) The pre-training data has a cutoff date of September 2025. The post-training data has a cutoff date of May 2026. NVIDIA Nemotron™ is a family of open models with open weights, training data, and recipes, delivering leading efficiency and accuracy for building specialized AI agents. NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 is a large language model (LLM) trained by NVIDIA. The model employs a hybrid Mixture-of-Experts architecture, utilizing interleaved Mamba-2 and MoE layers, along with select Attention layers. The Lightning 3.5 model is released alongside a number of speculative decoding methods for faster text generation. The model has 3B active parameters and 30B parameters in total.… Open weights other 17.8B parameters 1,048,576 tokens transformers [View model](https://savrn.com/models/nvidia-nemotron-3-5-lightning-30b-a3b-nvfp4) Model · Text generation ### [Qwen3.6-35B-A3B-NVFP4](https://savrn.com/models/qwen3-6-35b-a3b-nvfp4) [NVIDIA](https://savrn.com/model-publishers/nvidia) The NVIDIA Qwen3.6-35B-A3B-NVFP4 model is the quantized version of Alibaba's Qwen3.6-35B-A3B model, which is an auto-regressive language model that uses an optimized transformer architecture. For more information, please check here. The NVIDIA Qwen3.6-35B-A3B-NVFP4 model is quantized with Model Optimizer. This model is ready for commercial/non-commercial use. This model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA (Qwen3.6-35B-A3B) Model Card from Alibaba. Global Developers looking to take off-the-shelf, pre-quantized models for deployment in AI Agent systems, chatbots… Open weights apache-2.0 18.7B parameters 262,144 tokens Model Optimizer [View model](https://savrn.com/models/qwen3-6-35b-a3b-nvfp4) Model · Text generation ### [Qwen3-32B-NVFP4-W4A4](https://savrn.com/models/qwen3-32b-nvfp4-w4a4) [Cezar](https://savrn.com/model-publishers/ggamecrazy) NVFP4 W4A4 quantization of (revision 9216db5781bf21249d130ec9da846c4624c16137, BF16). It was made for a controlled NVFP4-vs-MXFP4 decode benchmark on NVIDIA B200 (fp4bench), not as a general-purpose release: the NVFP4 and MXFP4 checkpoints share the model, the tool, the recipe, the calibration data and the layer coverage, and differ only in the format. Recipe: QuantizationModifier(targets="Linear", scheme="NVFP4", ignore=["lmhead"]). The plain preset recipe, with no weight-rounding optimization (GPTQ, AutoRound and the like). anon8231489123/ShareGPTVicunaunfiltered at revision 192ab2185289094fc556ec8ce5ce1e8e587154ca, at most 1024 tokens each (33208 tokens in all, seed 3). The data sets… Open weights apache-2.0 19.1B parameters 40,960 tokens [View model](https://savrn.com/models/qwen3-32b-nvfp4-w4a4) Model · Text generation ### [K2-Horizon-MoVA-36B-A4B-NSLM-p4mx-q4v](https://savrn.com/models/k2-horizon-mova-36b-a4b-nslm-p4mx-q4v) [Matthew P Barnson](https://savrn.com/model-publishers/txgsync) An experimental NanoSeedLM (SeedLM) compression of IFM/K2-Horizon-MoVA-36B-A4B for Apple Silicon Macs with 32 GB of unified memory. SeedLM keeps each block of 8 weights as a 16-bit seed of a linear feedback shift register (LFSR), 4 coefficients and an exponent. The GPU makes the weights again from the seed at run time. The paper used an FPGA for this. This model uses Metal kernels on the Mac GPU. This is a research proof of concept. It is not a replacement for a calibrated quantization. - Six safetensors shards (22.76 GB in total, at most 4.5 GB each) and model.safetensors.index.json. The other files are config.json, the tokenizer files of the base model and nanoseedlmk2.py, the MLX loader.… Open weights apache-2.0 19.2B parameters 524,288 tokens mlx [View model](https://savrn.com/models/k2-horizon-mova-36b-a4b-nslm-p4mx-q4v) Model · Text generation ### [Ornith-1.5-35B-A3B-NVFP4](https://savrn.com/models/ornith-1-5-35b-a3b-nvfp4) [Ornith](https://savrn.com/model-publishers/ornith-ai) Chirp Chirp! We are introducing Ornith-1.5, a major step toward building foundation models through end-to-end self-improvement. Ornith-1.5 extends Ornith-1.0 (which was developed on top of Qwen3.5 and Gemma4 with additional continued pretraining, mid-training, and post-training) by expanding the self-improvement loop from scaffold and rollout optimization to jointly optimizing task generation, scaffold construction, and solution rollouts. Rather than relying on a fixed set of human-curated tasks and manually designed harnesses, Ornith-1.5 continuously generates new training tasks, discovers effective strategies for solving them, and improves the policy through reinforcement learning. For… Open weights mit 19.5B parameters 262,144 tokens transformers [View model](https://savrn.com/models/ornith-1-5-35b-a3b-nvfp4) ## NVIDIA [All models and datasets](https://savrn.com/model-publishers/nvidia) ## Versions - [6efb4a2a1c1f](https://savrn.com/models/nvidia-nemotron-3-nano-30b-a3b-nvfp4/versions/6efb4a2a1c1f) · current 2026-10-23 ## Explore More - [All text generation models](https://savrn.com/models/tasks/text-generation) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-19. - [NVIDIA website](https://build.nvidia.com/models) - [Hugging Face record](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4) - [How the hub is built](https://savrn.com/model-hub/methodology)