NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16: benchmark scores, price and ranking
NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 from nvidia ranks #44 of 743 LLMs on the LLMs Tiger aggregate benchmark score (53.1), based on 2 benchmark results captured up to 2026-10-13.
- Organisation
- nvidia
- Type
- Open-weight / local model
- Open weights
- Yes
- Aggregate rank
- #44 of 743 (score 53.1)
Across the 2 benchmarks where at least ten models are scored, NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 is in the top 10% on 1 and in the top half on 2.
NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| MMLU-Pro | 86.8% | #6 of 136 | HuggingFace MMLU-Pro leaderboard | 2026-10-13 |
| GPQA Diamond | 87.0% | #61 of 275 | HuggingFace GPQA Diamond leaderboard | 2026-10-13 |
About NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16
NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 is an open-weight model that can run on your own hardware from nvidia. No API price is listed for it yet. We hold 2 benchmark results for it across 1 area, and 1 is in the top tenth of its board.
Where NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 stands
On our aggregate score, NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 is #44 of 743 ranked models, in the top fifty. Its score is 53.1; the scale is centred on 50, so it sits 3.1 points above the typical model. That beats 94% of the ranked field. Among the 9 ranked nvidia models it is #2. The next one up is NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4 at 53.1. The next one down is NVIDIA-Nemotron-3-Super-120B-A12B-BF16 at 52.0.
NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16, benchmark by benchmark
On MMLU-Pro, NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 scores 86.8% and ranks #6 of 136. Only 5 models score higher, led by Darwin-180B-RSI at 88.1%, which is 1.3 points ahead. It covers 57-subject knowledge + reasoning. The board median is 55.3%, so NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 is 31.5 points above it. Source: HuggingFace MMLU-Pro leaderboard, captured 2026-10-13.
On GPQA Diamond, NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 reaches 87.0%. It ranks #61 of 275, ahead of 78% of the field, and 8.8 points behind GPT-6 Astra (max) at 95.8%. The results either side are grok-4-0709 (87.0%) and Darwin-27B-Opus (86.9%). It covers expert-level science questions. The board median is 71.7%, so NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 is 15.3 points above it. Source: HuggingFace GPQA Diamond leaderboard, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: strong, ahead of 87% of the field on average across 2 benchmarks. Its best board here is MMLU-Pro (#6 of 136) and its weakest is GPQA Diamond (#61 of 275).
Reading these numbers
The 2 scores on this page come from 2 sources: HuggingFace GPQA Diamond leaderboard (1), HuggingFace MMLU-Pro leaderboard (1). No official leaderboard supplied any of them, so treat them as third-party measurements. With 2 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #41 Grok 4.7 (xhigh) (53.1)
- #42 DeepSeek V4 Flash 0731 (max) (53.1)
- #43 o3 + gpt-4.1 (53.1)
- #45 Claude Code + Sonnet 5.5 (53.0)
- #46 Qwen3.5-122B-A10B (53.0)
- #47 Claude Mythos Preview (Early) (53.0)
More from nvidia
- NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4 (#40, 53.1)
- NVIDIA-Nemotron-3-Super-120B-A12B-BF16 (#83, 52.0)
- NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 (#114, 51.5)
- NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 (#117, 51.5)
- NVIDIA-Nemotron-3-Super-120B-A12B-FP8 (#181, 50.5)
- Nemotron-Cascade-2-30B-A3B (#194, 50.4)
Compare it on the live leaderboard → · How the aggregate score works