Nemotron TwoTower 30B-A3B Base

When to pick this model

• You're fine-tuning a custom instruct or reasoning model and want a lightweight MoE foundation
• You need 128K context without the VRAM cost of a dense 30B model
• You're building a research pipeline around NVIDIA's TwoTower MoE architecture
• Commercial fine-tuning under a permissive license matters

VRAM requirements by quantization

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 TwoTower 30B-A3B Base wants a 24 GB card at Q4_K_M (17 GB). Stepping up to Q8_0 nearly doubles the footprint to 32 GB, and unquantized FP16 weights take 60 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.

Without a GPU, Nemotron TwoTower 30B-A3B Base needs roughly 39 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 9 tokens/sec on entry-level GPUs, on the order of 14 tokens/sec on a mid-range card, and up to 22 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.

What hardware do you need

The table below matches Nemotron TwoTower 30B-A3B Base 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.

Which hardware should you buy to run Nemotron TwoTower 30B-A3B Base?

To run Nemotron TwoTower 30B-A3B Base locally at Q4, you need ~17 GB for Q4 weights alone. Hardware option to compare: GMKtec EVO-X2 64GB / 1TB (Ryzen AI Max+ 395). Leave memory for the system and context; verify inference-engine support. A mini PC does not provide CUDA or macOS/MLX.

Strengths

• MoE design (3B active of 30B total) keeps inference cost close to a much smaller dense model
• 128K native context out of the box
• Clean base for custom fine-tuning and alignment work
• NVIDIA Open Model License permits commercial use

Limitations

• Base checkpoint only — not instruction-tuned, unusable for chat until fine-tuned
• No official Ollama tag; requires manual HuggingFace setup
• Still needs ~17GB VRAM at Q4 quantization

Typical workloads

In our catalog grid, Nemotron TwoTower 30B-A3B Base is filed under Fine-Tuning Base, MoE Reasoning, Code Foundation — 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; multilingual workloads.

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: MoE, 30B total parameters / 3B active · BF16 weights · TwoTower Nemotron Labs architecture

Training: Base (non-instruct) model from the Nemotron Labs family. Native 128K token context.

Quick start

Install the runtime for your system: Windows, macOS or Linux. Check the exact model tag or GGUF quantization below; catalog IDs are not necessarily Ollama tags.

Start at 4096 tokens of context, then use ollama ps to check GPU/CPU placement. A default download may use a different quantization from the configurator’s memory estimate. Keep the free setup working before considering a kit.

# HuggingFace : nvidia/Nemotron-Labs-TwoTower-30B-A3B-Base-BF16

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.

or all the kits, for life — $49 →

Similar models worth comparing

Frequently asked questions

How much VRAM does Nemotron TwoTower 30B-A3B Base need?

At the recommended Q4_K_M quantization, Nemotron TwoTower 30B-A3B Base needs about 17 GB of VRAM. Q8_0 takes 32 GB, and unquantized FP16 weights take 60 GB.

Can Nemotron TwoTower 30B-A3B Base run without a GPU?

Yes — with roughly 39 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 TwoTower 30B-A3B Base support?

Nemotron TwoTower 30B-A3B Base supports a 125k-token context window (128,000 tokens).

Can I use Nemotron TwoTower 30B-A3B Base commercially?

Nemotron TwoTower 30B-A3B Base 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 TwoTower 30B-A3B Base on consumer hardware?

Our compatibility engine estimates on the order of 14 tokens/sec on a mid-range GPU and up to 22 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.

Which quantization of Nemotron TwoTower 30B-A3B Base should I download first?

Start with Q4_K_M (17 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. On a 24 GB card you can run up to Q5_K_M.