Nemotron-Labs-TwoTower-30B-A3B-Base-BF16

NVIDIAπŸ‡ΊπŸ‡Έ United States
active

Version History

base-bf16major

First release of Nemotron-Labs-TwoTower, a block-wise diffusion model built on the Nemotron-3-Nano-30B backbone. Uses dual-tower architecture to generate blocks of tokens in parallel, claiming 2.42Γ— speedup while retaining 98.7% of baseline quality.

Benchmark Scores

Full leaderboard β†’
BenchmarkScoreRankSource
MMLU-Pro60.9%70 of 76Hugging Face model card, Jul 2026 β†—

Rank is against every model in our database with a sourced score on that benchmark, each counted once with its best score.

Coverage

model releaseNVIDIA

NVIDIA releases Nemotron-Labs-TwoTower-30B: block-wise diffusion model claims 2.42Γ— faster generation at 98.7% baseline

NVIDIA released Nemotron-Labs-TwoTower-30B-A3B-Base-BF16, a block-wise diffusion language model that generates text by denoising blocks of tokens in parallel rather than sequentially. According to NVIDIA, the model achieves 2.42Γ— the wall-clock generation throughput of its autoregressive baseline while retaining 98.7% of aggregate benchmark quality.

2 min read