Skip to content
Local AI Wizard
Your machine
Theme

Nemotron 3.5 Lightning 30B A3B

NVIDIA · nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16

Q4 file 18.9 GB (Q4_0) · runs well from 32 GB · Weights license: OpenMDW 1.1

LiveBench · overall44.5 / 100#23 of 23 open models · 2026-06-25
Memory needed~23 GBQ4 weights + working room, before the machine's reserve
Parameters30 Bmanufacturer specification · 3B active
Context262,144 tokensPublished for this model

Quality

Quality · Overall

LiveBench 2026-06-25 · 0–100 · higher is better · this model outlined, #23 of 23

  • Color = publisher
  • Memory under review

Source: LiveBench 2026-06-25 · collected 2026-10-18 · open models without a score are not drawn (no score ≠ 0)

No result in LMArena 2026-10-22 Not in LMArena 2026-10-22

LiveBench categories

This model's mean per category · 0–100

42.9ReasoningReasoning45.8CodingCoding19.9Agentic CodingAgentic Coding66.5MathematicsMathematics43.1Data AnalysisData Analysis38.3LanguageLanguage54.9IFIF

Source: LiveBench 2026-06-25 · collected 2026-10-18 · open models without a score are not drawn (no score ≠ 0)

Reported by the publisher

The card has 4 evaluation tables, but none of their columns has been matched to this checkpoint yet (read 2026-10-25); no value is copied from another column. This stays an open gap. tables on the card, column not matched yet · read 2026-10-25

Run it

Run it

ollama run hf.co/ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF:Q4_0
  1. Install Ollama
  2. Run the command above
  3. Ask: Explain what a context window is, in two sentences.

Quality × memory

Quality × memory

LiveBench 2026-06-25 overall vs. Q4 planning memory (log scale) · 19 open models

  • This model
  • Color = publisher
40608010081632641282565121,0242,0484,096Memory needed · Q4 · model only (GB, log)LiveBenchSmaug AgenticQwen3.8 MaxDeepSeek V4 Pro · 0813Smaug MiniNemotron 3.5 Lightning 30B A3B

Memory = the model's own figure: parameters × 0.625 bytes + 20% working room, before any machine's reserve; with a chosen machine, the total there with its reserve (25% of memory, 2 to 8 GB; 2 GB per graphics card). An estimate, not measured VRAM. LiveBench 2026-06-25

Runs on

Fast memory holds it 42

The planning estimate fits the memory the model can use quickly: graphics card VRAM, or unified memory within the macOS GPU share.

Show all 42

Within memory, outside the fast part 51

Within the machine's memory, outside its fast part: on a graphics card the layers that do not fit run on the CPU from PC RAM (the listed RAM, or 32 GB assumed); on a Mac, above the macOS default GPU share. Not a speed figure.

Show all 51

One memory tier short 9

The same machines with the next memory size up would hold it in fast memory.

Show all 9

Nemotron 3.5 Lightning 30B A3B is a text model from NVIDIA with 30B parameters, 3B of them active per token (mixture of experts). It was released on August 11, 2026 and reads up to 262,144 tokens of context. Its Q4 weights plus working room come to about 23 GB before the machine's reserve; the smallest reference machines whose fast memory holds that are Mac mini M4 · 32 GB unified (~31 GB with its reserve) and RTX 5090 · 32 GB VRAM (~25 GB with its reserve). We collected 5 downloadable GGUF and MLX files from 2 publisher sources; the most downloaded GGUF source is ggml-org's. It scores 44.5 overall in LiveBench 2026-06-25.

Hardware profiles grouped by the Q4 planning estimate on each machine (dedicated GPUs reserve 2 GB; unified and system memory 25%, 2 to 8 GB): the fast memory holds it, it is within memory but outside the fast part (layers offloaded to PC RAM, or above the macOS GPU share), or the machine is one memory tier short. Each card opens its hardware profile. Not a measured fit or speed.

On the reference machines

Nemotron 3.5 Lightning 30B A3B on the reference machines
Where Q4 sits
Mac mini M2 · 8 GB unifiedLarger than its memoryNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~25 GBLarger than its memoryNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
Mac mini M4 · 16 GB unifiedLarger than its memoryNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~27 GBLarger than its memoryNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
Mac mini M4 · 24 GB unifiedLarger than its memoryNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~29 GBLarger than its memoryNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
Mac mini M4 · 32 GB unifiedFast memory holds it4bitMLX · 17.8 GB · needs ~30 GB262,144 tokens~31 GBFast memory holds it4bitMLX · 17.8 GB · needs ~30 GB262,144 tokens
Mac mini M5 Pro · 64 GB unifiedFast memory holds itQ8_0GGUF · 33.6 GB · needs ~49 GB262,144 tokens~31 GBFast memory holds itQ8_0GGUF · 33.6 GB · needs ~49 GB262,144 tokens
Mac Studio M5 Max · 128 GB unifiedFast memory holds itQ8_0GGUF · 33.6 GB · needs ~49 GB262,144 tokens~31 GBFast memory holds itQ8_0GGUF · 33.6 GB · needs ~49 GB262,144 tokens
Mac Studio · M5 Ultra · 36-core CPU / 80-core GPU · 512 GB unifiedFast memory holds itQ8_0GGUF · 33.6 GB · needs ~49 GB262,144 tokens~31 GBFast memory holds itQ8_0GGUF · 33.6 GB · needs ~49 GB262,144 tokens
RTX 4060 · 8 GB VRAMIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~25 GBIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
RTX 3060 · 12 GB VRAMIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~25 GBIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
RTX 5080 · 16 GB VRAMIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~25 GBIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
RTX 4090 · 24 GB VRAMIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~25 GBIn memory, outside the fast partNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
RTX 5090 · 32 GB VRAMFast memory holds itNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column~25 GBFast memory holds itNo file on the site's quant ladder fits the fast memory at 8,192 tokensNo file in the previous column
DGX H100 · 8 × 80 GBFast memory holds itQ8_0GGUF · 33.6 GB · needs ~57 GB262,144 tokens~39 GBFast memory holds itQ8_0GGUF · 33.6 GB · needs ~57 GB262,144 tokens

The file column uses the site's quant ladder (Q8_0 to Q3_K_M, MLX 8 to 4-bit on a Mac): the largest collected file whose size × 1.2 plus the context cache and the machine's reserve fits the fast memory. Planning estimates, not measured runs.

Plan another quantization or context

Consider it for

Compare Nemotron 3.5 Lightning 30B A3B using its publisher card and the LiveBench 2026-06-25 snapshot. This is quality evidence, not local speed.

Check first

The Q4 weights plus working room come to about 23 GB from 30B stated parameters, before the machine's reserve (Runs on adds it per machine). It is not a measured minimum; context, batch size, runtime and quantization change actual memory use.

Versions to download

Files you can download for this checkpoint. The section opens on the most downloaded collected GGUF source, ggml-org, at Q4_0. On a Mac it opens on the most downloaded MLX source. Sizes are the Hub's published bytes. What is Q4?

ggml-org

ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF · NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q4_0.gguf · 18.9 GB

Run this file

ollama run hf.co/ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF:Q4_0
llama-server -hf ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF:Q4_0
lms get https://huggingface.co/ggml-org/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF
  • mlx-lm: mlx-lm runs MLX repositories; pick the MLX version on a Mac

Quality scores on this page belong to the original checkpoint; these files are not scored or ranked. File sizes read 2026-10-23 · download counts read 2026-10-17 · commands: Hugging Face Hub docs: Use Ollama with any GGUF model · llama.cpp server README (-hf <user>/<model>[:quant]) · LM Studio docs: lms get with a Hugging Face URL · mlx-lm README (mlx_lm.generate --model)

Fine-tunes, adapters and other versions22 fine-tunes
Quantized versions
114
Fine-tunes
22
Adapters · merges
14 · 0

Browse them on the Hugging Face model tree · read 2026-10-23

Uncensored and abliterated versions8

Some community uploads remove the model's safety refusals. Quality may drop, the original publisher does not support them, claims about them are the uploader's, and the original license still applies. This site never recommends or ranks them.

8 derivative repositories have “uncensored” or “abliterated” in their name (read 2026-10-23).

Measured speed

No published measurement for this checkpoint

Tokens per second are shown only when a source publishes them with hardware, build and setup. They are never estimated.

How to run

The memory needed above is the Q4 load before a machine's reserve. Formula: parameters × 0.625 bytes, plus 20% working allowance; each machine then adds its system reserve: 25% of unified or system memory, 2 to 8 GB, or 2 GB per GPU on dedicated cards. This is neither measured VRAM nor checkpoint download size.

Q4 planning estimate only. Check checkpoint format, runtime support, KV cache, context, batch size and offload before choosing hardware.

Evidence level: Generic estimate only; no execution reproduced by this site.

Publisher sampling defaults
temperature 1 · top_p 0.95generation_config.json · read 2026-10-23source
Chat template markers
tool calls · thinkingmarkers found in the template, not a test of the feature · read 2026-10-23source
Before downloading
  1. Check the exact checkpoint and license linked here.
  2. Confirm a compatible runtime and precision or quantization for your platform.
  3. Allow for context, batch size and any extra media components beyond the weights.

Technical details

What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.

Task and modality
language and reasoning · text
Count and basis
30B · manufacturer specification · 3B activenvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 model card (Hugging Face) · read 2026-10-17source
Context
262,144 tokens published for this modelnvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 config.json (Hugging Face) · read 2026-10-17source
Architecture
MoEnvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 model card (Hugging Face) · read 2026-10-17source
Released
nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 model card (Hugging Face) · read 2026-10-17source
Runtime and precision
No local execution verified.
Languages
en, es, fr, de, it, jaModel card metadata · read 2026-10-25source
Commercial use
Allowed OpenMDW 1.1 text (nvidia/NV-Reason-CT LICENSE) · read 2026-10-18license text
Every stored field12 fields
Revision checked
a9904d24bcc1d289a1950fa9d2b978c47cf903b9
Architecture
NemotronHForCausalLM
Library
transformers
Hub task tag
text-generation
Formats
safetensors
Access
Open download
Publisher kind
Official
Repository created
2026-08-01creation is not the model's release
Repository updated
2026-08-24a commit is not a release date
Hub downloads (30 days) · likes
861,461 · 251this repository only · read 2026-10-25
Read from the Hub page
transformers class, task tags, license nameHub overview and file listing · read 2026-10-18
Catalog state
Hub config or weights found; no inventory benchmark link

Hybrid Mamba-2, MoE and attention model. The linked repository is the BF16 reference checkpoint, which NVIDIA points to for customization; it points to a separate NVFP4 release for optimized inference. The generic Q4 estimate is not a measured requirement.

Sources

Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.

Catalog reviewed: 2026-10-17. Benchmark release: 2026-06-25; snapshot collected: 2026-10-18.