88B (MoE-derived optimized architecture)
NVIDIA Open Model License
Deployment-optimized reasoning and chat model produced with the Puzzle post-training NAS framework to improve inference efficiency under serving constraints. Official model-card claims report substantial throughput gains in both long-context and short-context H100 deployments, while preserving or slightly improving reasoning accuracy across effort settings. The release describes heterogeneous MoE expert pruning, selective window attention for KV-cache efficiency, and RoPE/YaRN scaling adjustments to support stable long-context behavior (128K). Post-architecture optimization includes knowledge distillation and RL refinement with effort-controlled reasoning modes (low/medium/high) for configurable cost-quality trade-offs. Card notes this checkpoint is intended for production-grade, cost-efficient reasoning workloads and includes open deployment guidance and dataset disclosure sections.