Nemotron 3 Nano 30B-A3B
- Made by
- NVIDIA
- Released
- Announced in
- NVIDIA Debuts Nemotron 3 Family of Open Modelsnvidianews.nvidia.com
- Event
- Nemotron 3
- API record
- /api/v1/models/nemotron-3-nano-30b-a3b
Benchmark results
| Benchmark | Score | Source check |
|---|---|---|
| Humanity's Last ExamNo tools | 10.2% | CorroboratedThe same number is printed in 1 other lab document: 1 |
| in Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning | 10.57% | CorroboratedThe same number is printed in 1 other lab document: 1 |
| later re-run; text-only subset · reported later, in nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 model card (revision of 2026-08-11) | 10.89% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| Humanity's Last ExamWith tools | 15.5% | |
| GPQA Diamondno tools | 73.04% | |
| later re-run; no tools · reported later, in nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 model card (revision of 2026-08-11) | 74.05% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| AIME 2025No tools | 89.1% | |
| AIME 2025With tools | 99.2% | |
| MMLU-Pro | 78.3% | |
| LiveCodeBenchLiveCodeBench v6 | 68.3% | |
| LiveCodeBench v6 · in NVIDIA Nemotron 3 Family of Models | 68.2% | CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced. |
| τ²-benchTelecom | 42.2% | |
| τ²-benchRetail | 56.9% | |
| τ²-benchAirline | 48.0% |
Sources
Reported by NVIDIA
- nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8 model card (launch-day revision)huggingface.co · 15 Dec 2025
Our capture of the source - NVIDIA Nemotron 3 Family of Modelsresearch.nvidia.com · 15 Dec 2025
Original image - Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoningresearch.nvidia.com · 15 Dec 2025
Our capture of page 22 - nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 model card (revision of 2026-08-11)huggingface.co · 11 Aug 2026
Our capture of the source
