Llama-3.3-nemotron-super-49B-v1.5 (Reasoning): benchmark results

Provider: NVIDIA. Released 2025-07-25. Access: Open.

Aggregate Rating 1544 ± 12, rank #915 of 2201 rated models, from 96 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
NVIDIA Nemotron 49B v1.5 Reproduction - aime24 - majority@16 - symbolic_correct93.33Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - aime24 - pass@16 - symbolic_correct93.33Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - aime24 - pass@1[avg-of-16] - symbolic_correct88.54Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - aime25 - majority@16 - symbolic_correct93.33Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - aime25 - pass@16 - symbolic_correct93.33Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - aime25 - pass@1[avg-of-16] - symbolic_correct84.38Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - bfcl_v3 - live_ast82.68Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - bfcl_v3 - live_relevance77.78Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - bfcl_v3 - non_live_ast88.58Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - bfcl_v3 - overall_accuracy72.64Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - bfcl_v3 - overall_live83.34Accuracy (%)100
NVIDIA Nemotron 49B v1.5 Reproduction - bfcl_v3 - overall_multi_turn46.38Accuracy (%)100

Interactive version: theaggregate.ai/model?slug=llama-3-3-nemotron-super-49b-v1-5-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-12.