A high-performance Mixture of Experts (MoE) model developed by NVIDIA. It features 550 billion parameters and is optimized for GPUs.
Parameters
Undisclosed
Context Window
License
FREE-COMMERCIAL
Release Date
2026-06-04
Benchmark Performance
AA Intelligence Index
—
LMArena Elo
—
HLE
—
ARC-AGI-2
—
SWE-bench Verified
—
GPQA Diamond
—
MMLU-Pro
—
LiveCodeBench
—
AIME 2025
—
MATH-500
—
API Pricing
API pricing for this model is not yet available