BlogLLM Benchmark Scores as Published, Side by Side
Nemotron 3 Nano 30B-A3B: Benchmark Scores as Published
Every result the catalog holds for Nemotron 3 Nano 30B-A3B: 40 tests in 11 areas, from 3 pages. 36 printed by NVIDIA itself, the rest by the company named beside each number. Each number is as printed; nothing is ranked or rescored.
Developer NVIDIAReleased 2025-12-15Context window 1MMax output 1MInput TextOutput TextWeights Open (NVIDIA Nemotron Open Model License)Data version 2026103001, as of 2026-10-30
Knowledge Work
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| GDPval-AA v2 | 473 | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
Agentic Coding
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| SWE-bench Verified | 34.08% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
| Terminal-Bench 2.1 | 8.29% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
| SWE-bench Multilingual | 14.07% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
| Terminal-Bench Hard | 8.5% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| SWE-bench (split not stated) | 38.8% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Harness: OpenHands |
Science and Math
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPQA Diamond | 74.05% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | Tools: none |
| AIME 2025 | 99.2% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Tools: with tools |
| 89.1% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Tools: none | |
| SciCode | 30.08% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19; printed on 1 more page | - |
| GPQA | 73% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Tools: none |
| 75% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Tools: with tools | |
| MiniF2F (pass@1) | 50% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| MiniF2F (pass@32) | 79.9% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
Long Context
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| AA-LCR | 32.75% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19; printed on 1 more page | - |
| RULER-100 at 1M | 86.3% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Context: 1048576 |
| RULER-100 at 256K | 92.9% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Context: 262144 |
| RULER-100 at 512K | 91.3% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Context: 524288 |
General Knowledge and Reasoning
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Humanity's Last Exam (no tools) | 10.89% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19; printed on 1 more page | Tools: none |
| MMLU-Pro | 78.46% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19; printed on 1 more page | - |
| Humanity's Last Exam (with tools) | 15.5% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Tools: with tools |
| AA-Omniscience Accuracy | 20.15% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
| FACTS Grounding (0 to 1) | 0.48 | AI21 Labs, AI21 Jamba2 Mini model cardchecked 2026-10-15 | Effort: non-thinking |
Code Generation
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| LiveCodeBench v6 | 68.3% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
Tool Use and Workflows
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| τ³-bench Banking | 7.01% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
| τ²-bench Telecom | 42.2% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| τ²-bench Retail | 56.9% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| τ²-bench Airline | 48% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| PinchBench | 66.11% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
| BFCL v4 | 53.8% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| τ²-bench (average of 3 domains) | 49% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
Web Search and Research
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| BrowseComp | 13.74% | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19 | - |
Instruction Following
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| IFBench | 72.17 | NVIDIA, NVIDIA Nemotron 3.5 Lightning 30B-A3B model cardchecked 2026-10-19; printed on 2 more pages | - |
| MultiChallenge | 38.5% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| IFEval | 88% | AI21 Labs, AI21 Jamba2 Mini model cardchecked 2026-10-15 | Effort: non-thinking |
| COLLIE | 30% | AI21 Labs, AI21 Jamba2 Mini model cardchecked 2026-10-15 | Effort: non-thinking |
| Enterprise Reliability Score (average, 0 to 1) | 0.5 | AI21 Labs, AI21 Jamba2 Mini model cardchecked 2026-10-15 | Effort: non-thinking |
Multilingual
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| MMLU-ProX | 59.5% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| WMT24++ (XCOMET-XXL) | 86.2 | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
Preference and Composite Rankings
| Test | Value | Printed by | Setup as printed |
|---|---|---|---|
| Arena-Hard v2 (average) | 67.7% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| Arena-Hard v2 (creative writing) | 63.2% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| Arena-Hard v2 (hard prompt) | 72.1% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
More from NVIDIA
Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.