NVIDIA Unveils Nemotron 3.5 Lightning and NeMo Switchyard Router
NVIDIA released Nemotron 3.5 Lightning and NeMo Switchyard on August 11, targeting efficient execution for AI agents from edge to cloud. The model uses a hybrid Mamba-Transformer MoE architecture with 30B total parameters (31.6B measured) and 3B active (3.6B measured), supporting 1M context and measured 670 tokens/s output, leading in agentic coding tasks. Weights are open-sourced on Hugging Face, runnable on single GPU, with NVFP4 version topping HF trends. NeMo Switchyard dynamically routes based on request signals, agent state, tool results, and model capabilities, cost, latency.
Confirmed
- Architecture and performance: hybrid Mamba-Transformer MoE, 30B total (31.6B measured), 3B active (3.6B measured), 1M context, NVFP4 and BF16, 670 tokens/s measured.
- Open source and deployment: weights on Hugging Face, free, single GPU, DGX Spark demo, Baseten and Ollama support, NVFP4 topped HF trends.
- Router: NeMo Switchyard open source, multi-dimensional dynamic routing.
Why it matters
Targets AI agent execution layer pain points, reducing high-frequency call cost and latency via ultra-low active parameter MoE, and with Switchyard dynamic routing, improves compute allocation for complex tasks, promoting always-on agents.
2026-08-11 ~ 2026-08-13 · 55 related posts
- Episode 1: NVIDIA Unveils Nemotron 3.5 Lightning and NeMo Switchyard Router(2026-08-11, 55 posts)
- Episode 2: Fine-tuned NVIDIA Nemotron 3.5 Outperforms Claude Opus(2026-08-12, 4 posts)
- Episode 3: Nvidia Switchyard Router Cuts Agent Costs by 59%(2026-08-12, 3 posts)
- Episode 4: NVIDIA Nemotron 3.5 Excels in Scientific Post-Training(2026-08-12, 2 posts)
- Episode 5: Meta Open-Sources 30B Local Agent Model Muse Glimmer(2026-08-12, 2 posts)
- Episode 6: Meta and NVIDIA Launch 30B Open-Source Agent Models on the Same Day(2026-08-12, 3 posts)
- Episode 7: NVIDIA Open Sources NeMo Switchyard for Dynamic Agent Routing(2026-08-15, 2 posts)
- Episode 8: NVIDIA Unveils Nemotron 3.5 Lightning and NeMo Switchyard(2026-08-19, 3 posts)
Primary sources
- NVIDIA Releases Nemotron-3.5-30B-A3B: An Open MoE Model — coder543 ·
- NVIDIA Launches Nemotron Lightning: A Super Fast MoE Model for Long-Running Agents — Sam Witteveen ·
- NVIDIA Open-Sources NeMo Switchyard for Dynamic AI Agent Routing — Sam Witteveen ·
- NVIDIA Walks Through Deploying Nemotron 3.5 Lightning on DGX Spark for Agents — NVIDIA Developer · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning Model — NVIDIAAI · 2026-08-11
- [source] NVIDIA Open-Sources NeMo Switchyard for Dynamic AI Agent Routing — Sam Witteveen · 2026-08-11
- NVIDIA Unveils Nemotron 3.5 Lightning, Built for Long-Running Agents — NVIDIA Developer · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning: Ultra-Fast Open Model for Agents — ArtificialAnlys · 2026-08-11
- Nemotron 3.5 Lightning Tested: 670 tokens/s Speedster for Agents — ArtificialAnlys · 2026-08-11
- NVIDIA Nemotron 3.5 Lightning Hits DeepInfra with 1M Token Context — gharik · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning and NeMo Switchyard for Agentic AI — nordicinst · 2026-08-11
- Video: Building a Local LLM Routing Layer with NVIDIA SwitchYard — Prompt Engineering · 2026-08-11
- Nemotron 3.5 Lightning Benchmarks: 16x Throughput Scale with No Latency Hit — rhythmrg · 2026-08-11
- [source] NVIDIA Releases Nemotron-3.5-30B-A3B: An Open MoE Model — coder543 · 2026-08-11
- [source] NVIDIA Launches Nemotron Lightning: A Super Fast MoE Model for Long-Running Agents — Sam Witteveen · 2026-08-11
- NVIDIA Nemotron 3.5 Lightning Weights Now on Hugging Face — MaziyarPanahi · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning: 4x Faster Open Model for Agents — kuchaev · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning Optimized for Agentic Execution — NVIDIAAI · 2026-08-11
- NVIDIA Open-Sources Nemotron 3.5 Lightning for High-Volume Agentic Execution — NVIDIAAI · 2026-08-11
- Nvidia Releases Nemotron 3.5 Lightning Open-Source Model — Polymarket · 2026-08-11
- NVIDIA releases Nemotron 3.5 Lightning: 30B LatentMoE, 3B active, 1M context — huggingface · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning: 30B Params with 3B Active, Built for Agent Execution — heyshrutimishra · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning with 1M Token Context — mervenoyann · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning: 30B MoE Built for Always-On Agents — ccerrato147 · 2026-08-11
- Nemotron 3.5 Specs Revealed: 3.6B Active Params Matches Larger Models — thisguyknowsai · 2026-08-11
- NVIDIA Launches Nemotron 3.5 Lightning for Long-Running Agents — rohanpaul_ai · 2026-08-11
- Hands-on: NVIDIA Nemotron 3.5 Hits 250+ tokens/sec on 30B MoE — Arindam_1729 · 2026-08-11
- NVIDIA Nemotron 3.5 Lightning Goes Live on CoreWeave Serverless — wandb · 2026-08-12
- Ollama Adds Support for NVIDIA's Nemotron 3.5 Lightning Model — ollama · 2026-08-12
- NVIDIA Launches Nemotron 3.5 Lightning for Local AI Agents — ollama · 2026-08-12
- OpenHands Integrates NVIDIA Nemotron to Advance Open, Composable Coding Agents — rajistics · 2026-08-12
- NVIDIA's Nemotron 3.5 Lightning 30B Model Hits Hugging Face Trending — nvidia · 2026-08-12
- NVIDIA Launches Nemotron 3.5 Lightning and NeMo Switchyard for Local Agentic AI — sudoraohacker · 2026-08-12
- NVIDIA Open-Sources Nemotron Agentic Coding RL Dataset — NVIDIAAI · 2026-08-12
- NVIDIA Releases Open-Source 30B Nemotron Model Optimized for AI Agents — DeryaTR_ · 2026-08-12
- NVIDIA's 30B MoE Model Lands on Perplexity Agent API — denisyarats · 2026-08-12
17 near-duplicate retellings: NVIDIA Developer · MaziyarPanahi · nvidia · NVIDIAAI · baseten · baseten · kuchaev · tw_killian · dr_alphalyrae · baseten · ccerrato147 · solyarisoftware · ziqiao_ma · nvidia · willccbb · rohanpaul_ai · ctnzr