Nemotron 3 Super 12B
By NVIDIA · United States
Updated 2026-08-28
Overview
A dense 12B distillation of NVIDIA's larger Nemotron 3 Super MoE, compact and fast enough for consumer GPUs while retaining strong reasoning and code performance.
When to pick this model
- Reasoning and code assistance on a single consumer GPU
- Teams wanting NVIDIA's Nemotron lineage without datacenter hardware
- Agentic workflows needing a fast, dense mid-size model
- Local deployment on 12GB-class GPUs or Apple Silicon
VRAM requirements by quantization
| Quantization | VRAM required |
|---|---|
| Q4_K_M (recommended) | 7 GB |
| Q5_K_M | 9 GB |
| Q8_0 | 13 GB |
| FP16 (no quantization) | 24 GB |
VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.
In practice, Nemotron 3 Super 12B fits an 8 GB consumer card at Q4_K_M (7 GB). Stepping up to Q8_0 nearly doubles the footprint to 13 GB, and unquantized FP16 weights take 24 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.
Without a GPU, Nemotron 3 Super 12B needs roughly 16 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 8 tokens/sec on entry-level GPUs, on the order of 25 tokens/sec on a mid-range card, and up to 65 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.
What hardware do you need
The table below matches Nemotron 3 Super 12B to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.
| GPU memory | Example cards | Best fit for Nemotron 3 Super 12B |
|---|---|---|
| 8 GB | RTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GB | Q4_K_M (7 GB used) |
| 12 GB | RTX 5070, RTX 5070 Ti Laptop, RTX 4080 Laptop | Q5_K_M (9 GB used) |
| 16 GB | RTX 5080, RTX 4080 Super, Radeon RX 9070 XT | Q8_0 (13 GB used) |
| 24 GB | RTX 4090, Radeon RX 7900 XTX, RTX 5090 Laptop | FP16 (24 GB used) |
| 32 GB | RTX 5090 | FP16 (24 GB used) |
Which GPU should you buy to run Nemotron 3 Super 12B?
To run Nemotron 3 Super 12B locally at Q4, you need ~7 GB of VRAM. The best value for this is a RTX 5060 (8 GB VRAM).
As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.
Strengths
- Compact dense 12B footprint (7GB VRAM at Q4)
- Strong reasoning and code performance
- Runs on a 12GB RTX 3060-class GPU or Apple Silicon
- NVIDIA Open Model License
Limitations
- Gated on Hugging Face (click-through access required)
- Restrictive commercial terms under NVIDIA's Open Model License
- Native context window not publicly confirmed
Typical workloads
In our catalog grid, Nemotron 3 Super 12B is filed under Reasoning, Code, Agents — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: code generation and review (pair it with an editor integration like Continue.dev or Cline); multi-step reasoning and math-flavoured tasks.
The 125k-token context window covers long chats and mid-sized documents, though very large retrieval workloads will need chunking. It ships under the NVIDIA Open Model License license — commercial use is generally possible but read the specific terms before embedding it in a product.
Architecture & training
Architecture: Dense Transformer · 12B parameters · compact variant of the Nemotron 3 Super family
Training: NVIDIA's Nemotron 3 Super family. Dense 12B variant distilled from the 120B frontier MoE model.
A distilled, consumer-GPU-friendly 12B that inherits Nemotron 3 Super's reasoning strength at a fraction of the size.
Quick start
ollama pull nemotron-3-superOr use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.
Similar models worth comparing
Frequently asked questions
How much VRAM does Nemotron 3 Super 12B need?
At the recommended Q4_K_M quantization, Nemotron 3 Super 12B needs about 7 GB of VRAM. Q8_0 takes 13 GB, and unquantized FP16 weights take 24 GB.
Can Nemotron 3 Super 12B run without a GPU?
Yes — with roughly 16 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.
What context window does Nemotron 3 Super 12B support?
Nemotron 3 Super 12B supports a 125k-token context window (128,000 tokens).
Can I use Nemotron 3 Super 12B commercially?
Nemotron 3 Super 12B ships under the NVIDIA Open Model License license. Commercial use is generally permitted subject to its terms — review the license text before shipping a product.
How fast is Nemotron 3 Super 12B on consumer hardware?
Our compatibility engine estimates on the order of 25 tokens/sec on a mid-range GPU and up to 65 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.
Which quantization of Nemotron 3 Super 12B should I download first?
Start with Q4_K_M (7 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. It already fits an 8 GB card at Q4_K_M.