NVIDIA AI & NIMOpen weights
Params
550B MoE (55B active)Context
1.0MSpeed
—Release
Apr 2025NVIDIA AI & NIM
Leading accelerated computing architecture creator behind the Nemotron foundation series, NeMo framework, and NVIDIA NIM inference microservices.
Models by NVIDIA AI & NIM (11)
NVIDIA AI & NIMCloud API
Params
UndisclosedContext
1.0MSpeed
—Release
Aug 2026NVIDIA AI & NIMOpen weights
Params
70.6B DenseContext
128kSpeed
—Release
Oct 2024NVIDIA AI & NIMOpen weights
Params
UndisclosedContext
262kSpeed
—Release
Jun 2026NVIDIA AI & NIMOpen weights
Params
120B MoE (12B active)Context
262kSpeed
—Release
Mar 2025NVIDIA AI & NIMOpen weights
Params
340B DenseContext
4kSpeed
—Release
Jun 2024NVIDIA AI & NIMOpen weights
Params
340B Dense RewardContext
4kSpeed
—Release
Jun 2024NVIDIA AI & NIMOpen weights
Params
UndisclosedContext
262kSpeed
—Release
Jun 2026NVIDIA AI & NIMOpen weights
Params
UndisclosedContext
1.0MSpeed
—Release
Jun 2026NVIDIA AI & NIMOpen weights
Params
UndisclosedContext
131kSpeed
—Release
Jun 2026NVIDIA AI & NIMOpen weights
Params
UndisclosedContext
262kSpeed
—Release
Jun 2026