Nemotron 3 Super: benchmark scores, price and ranking
Nemotron 3 Super from Nvidia ranks #599 of 743 LLMs on the LLMs Tiger aggregate benchmark score (45.3), based on 4 benchmark results captured up to 2026-10-13.
- Organisation
- Nvidia
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2026-03-11 (source: epoch)
- Cheapest API price
- $0.080 in / $0.450 out per million tokens (via openrouter, observed 2026-10-13T20:47Z)
- Aggregate rank
- #599 of 743 (score 45.3)
Relative to other models, Nemotron 3 Super is strongest on CritPt (#56 of 115) and weakest on ALE-Bench (#76 of 79).
Across the 4 benchmarks where at least ten models are scored, Nemotron 3 Super is in the top 10% on 0 and in the top half on 1.
Its cheapest listed input price, $0.080 per million tokens, is lower than that of 85% of the 427 models with a listed price.
Nemotron 3 Super benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| CritPt | 3.1% | #56 of 115 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WeirdML | 38.0% | #69 of 109 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SciCode | 36.0% | #77 of 106 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ALE-Bench | 214 | #76 of 79 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Nemotron 3 Super
Nemotron 3 Super is an open-weight model that can run on your own hardware from Nvidia. It was released on 2026-03-11, according to Epoch AI. The cheapest API listing we track charges $0.080 per million input tokens and $0.450 per million output tokens, through openrouter. We hold 4 benchmark results for it across 3 areas, and none is in the top tenth of its board.
Where Nemotron 3 Super stands
On our aggregate score, Nemotron 3 Super is #599 of 743 ranked models, in the lower half. Its score is 45.3; the scale is centred on 50, so it sits 4.7 points below the typical model. That beats 19% of the ranked field. Among the 2 ranked Nvidia models it is #2. The next one up is nemotron-3-ultra at 47.0.
Nemotron 3 Super, benchmark by benchmark
CritPt puts Nemotron 3 Super at 3.1%, #56 of 115. That beats 51% of the models tested, but 29.1 points short of GPT-5.6 Sol (max) (32.3%). The results either side are Kimi K2.5 (3.1%) and nemotron-3-ultra (3.1%). It covers unpublished research-level physics problems. The board median is 2.9%, so Nemotron 3 Super is 0.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Nemotron 3 Super lands below the middle on WeirdML: 38.0%, #69 of 109, with 62% of the field scoring higher. GPT-6 Astra (max) leads at 93.3%; the closest result above is GPT-5 nano (high) at 38.1%. It covers unusual machine-learning tasks. The board median is 41.2%, so Nemotron 3 Super is 3.2 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Nemotron 3 Super lands below the middle on SciCode: 36.0%, #77 of 106, with 72% of the field scoring higher. Claude Opus 5.5 (max) leads at 66.9%; the closest result above is Trinity Large Thinking at 36.1%. It covers scientific research code. The board median is 44.3%, so Nemotron 3 Super is 8.3 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
ALE-Bench is a weak spot. Nemotron 3 Super scores 214, #76 of 79, behind 95% of the models measured. The leader, GPT-6 Astra (max), reaches 2,951. It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 764, so Nemotron 3 Super is 550 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Software engineering: weak, ahead of 16% of the field on average across 2 benchmarks. Its best board here is SciCode (#77 of 106) and its weakest is ALE-Bench (#76 of 79).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Nemotron 3 Super costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on Qwen3-30B-A3B ($0.090) and DeepSeek-R1-Distill-Qwen-1.5B ($0.090).
Release history
The next one was nemotron-3-ultra on 2026-06-04, at 47.0.
Reading these numbers
The 4 scores on this page come from 1 source: Epoch AI Benchmarking Hub (4). No official leaderboard supplied any of them, so treat them as third-party measurements. With 4 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #596 Llama-2-13b-chat (45.3)
- #597 WizardLM-2-8x22B (45.3)
- #598 c4ai-command-r7b-12-2024 (45.3)
- #600 Llama-2-70b-chat (45.3)
- #601 Mistral-Nemo-Instruct-2407 (45.2)
- #602 claude-haiku-4-5-20251001 (45.2)
More from Nvidia
- nemotron-3-ultra (#473, 47.0)
Compare it on the live leaderboard → · How the aggregate score works