Llama Nemotron Nano 4B v1.1Open Weight

Released

Llama Nemotron Nano 4B v1.1 is an AI model released by NVIDIA on Tuesday, May 20 2025, 43 days after Llama Nemotron Ultra 253B. It is an open-weight model — the trained weights are available to download and run. It is a 4B parameterThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. model with a 128k token context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider.. Benchmark results (shown below) cover IFEval, MATH-500, AIME 2025, and GPQA Diamond.

Benchmarks

Agentic & tool use

IFEvalAgentic tasks — following verifiable instructions in generated responses
75.5%
#21 of 24

Reasoning & science

MATH-500Reasoning — accuracy on the 500-problem MATH subset
96.2%
#16 of 28
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools
46.3%
#47 of 47
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better.
55.1%
#128 of 145

Compare Llama Nemotron Nano 4B v1.1 with

Llama Nemotron Nano 4B v1.1

About

NVIDIA released Llama Nemotron Nano 4B v1.1 in May 2025 as a smaller reasoning model for local use. NVIDIA post-trained its Minitron 4B base checkpoint, which it had pruned from Meta’s Llama 3.1 8B.

NVIDIA trained the model for mathematics, code and assistant tasks. Its reasoning mode increased its reported GPQA Diamond score from 33.8% to 55.1%, illustrating the difference that the inference setting made on the same weights.

Frequently asked questions

Llama Nemotron Nano 4B v1.1 was released by NVIDIA on Tuesday, May 20 2025.

All NVIDIA releases

20 tracked