--- title: The Fastest Nemotron Yet: Embracing NVIDIA Nemotron 3.5... description: NVIDIA Nemotron 3.5 Lightning is a 30B Mixture-of-Experts model with only 3B active parameters, distilled from the Nemotron 3 Ultra frontier model. It supports... canonical: https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The Fastest Nemotron Yet: Embracing NVIDIA Nemotron 3.5 Lightning in Kilo | daily.dev og:description: NVIDIA Nemotron 3.5 Lightning is a 30B Mixture-of-Experts model with only 3B active parameters, distilled from the Nemotron 3 Ultra frontier model. It supports... og:url: https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll og:image: https://api.daily.dev/og/posts/WcpGEC5lL.png og:image:alt: The Fastest Nemotron Yet: Embracing NVIDIA Nemotron 3.5 Lightning in Kilo og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # The Fastest Nemotron Yet: Embracing NVIDIA Nemotron 3.5 Lightning in Kilo **[Kilo Blog](https://daily.dev/sources/kilo-ai-blog)** · 4 min read · 4 upvotes · 2 comments ## Summary NVIDIA Nemotron 3.5 Lightning is a 30B Mixture-of-Experts model with only 3B active parameters, distilled from the Nemotron 3 Ultra frontier model. It supports context windows up to 1M tokens and claims up to 4x higher throughput than competing models. Kilo's engineers tested it in production agentic workflows and found it excels as a sub-agent executor when given strict, well-defined instructions. The Thinking variant achieved ~73% exact completion on git benchmarks and a perfect 16/16 on deterministic git operations. The model has a notably opinionated personality, making it less suited for open-ended planning but highly effective in structured multi-agent pipelines. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is NVIDIA Nemotron 3.5 Lightning and how big is it? NVIDIA Nemotron 3.5 Lightning is a customizable open model built as a 30B parameter Mixture-of-Experts architecture with only 3B active parameters, distilled from NVIDIA's frontier Nemotron 3 Ultra model. It supports context windows up to 1M tokens and offers up to 4x higher throughput than some competing models, making it optimized for always-on, always-running agentic workloads. _Developers picking a fast open model for agent workloads can compare benchmarks like these on daily.dev._ ### What are the Instant and Thinking variants of Nemotron 3.5 Lightning good for? The Instant variant handles low-risk, deterministic git operations well, achieving a 100% success rate on tagging plus strong results on undo/reset recovery, commit construction, and remote handling. The Thinking variant achieved a nearly 73% exact completion rate on a git benchmark and scored perfectly (16/16) on deterministic operations like repository inspection, revert, stash, and undo recovery. _Teams choosing between model variants for git-heavy agent tasks can track results like these on daily.dev._ ### Is Nemotron 3.5 Lightning good for high-level agent planning? Not really. It has a highly opinionated, spirited personality that makes it a poor fit for unconstrained, high-level planning, but it excels as a specialized executor when given strict guidelines and bounded instructions, such as within an orchestrator-executor multi-agent setup where it reliably completes clearly defined sub-tasks without drift. _Developers designing multi-agent orchestration setups can follow model behavior notes like these on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@kartiknvj** · 1 upvotes > The finding that it shines as a sub-agent executor but not as an open-ended planner is more useful than the throughput number, and it matches how I have been splitting roles: a strong planner up top, cheap fast models doing the constrained steps underneath. It also means you have to evaluate it on the executor task specifically, since a planning benchmark would make it look worse than it is. Where did the 4x throughput actually land once the sub-agents were doing real tool calls? **@agustinbarrientos** · 0 upvotes > I'd keep this one on git chores and give planning to a model with fewer opinions. ## Similar posts on daily.dev - [NVIDIA Nemotron 3.5 Lightning · Ollama Blog](https://daily.dev/posts/nvidia-nemotron-3-5-lightning-ollama-blog-uy1kba1ur) · Ollama · 3 upvotes · 0 comments - [NVIDIA Nemotron 3 Ultra is Here – and it’s Free to Use in Kilo](https://daily.dev/posts/nvidia-nemotron-3-ultra-is-here-and-it-s-free-to-use-in-kilo-qmednaxh7) · Kilo Blog · 1 upvotes · 0 comments - [Run Highly Efficient and Accurate Multi-Agent AI with NVIDIA Nemotron 3 Super Using vLLM](https://daily.dev/posts/run-highly-efficient-and-accurate-multi-agent-ai-with-nvidia-nemotron-3-super-using-vllm-cpvrpiesj) · vLLM · 1 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents), [#mixture-of-experts](https://daily.dev/tags/mixture-of-experts) [View this post on daily.dev](https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"The Fastest Nemotron Yet: Embracing NVIDIA Nemotron 3.5 Lightning in Kilo","url":"https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll"},"datePublished":"2026-08-11T16:23:53.837Z","dateModified":"2026-09-13T20:20:50.822Z","description":"NVIDIA Nemotron 3.5 Lightning is a 30B Mixture-of-Experts model with only 3B active parameters, distilled from the Nemotron 3 Ultra frontier model. It supports...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/658aa3e22bfbf4e3a6504604a1ef7b9d?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/658aa3e22bfbf4e3a6504604a1ef7b9d?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Kilo Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Kilo Blog","logo":"https://media.daily.dev/image/upload/s--x3pmrf8D--/f_auto,q_auto/v1774964071/logos/kilo-ai-blog?_a=BAMAMiWQ0","url":"https://daily.dev/sources/kilo-ai-blog"},"commentCount":2,"discussionUrl":"https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":2}],"keywords":"llm,ai-agents,mixture-of-experts","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Kilo Blog","item":"https://daily.dev/sources/kilo-ai-blog"},{"@type":"ListItem","position":3,"name":"The Fastest Nemotron Yet: Embracing NVIDIA Nemotron 3.5 Lightning in Kilo"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll","comment":[{"@type":"Comment","text":"The finding that it shines as a sub-agent executor but not as an open-ended planner is more useful than the throughput number, and it matches how I have been splitting roles: a strong planner up top, cheap fast models doing the constrained steps underneath. It also means you have to evaluate it on the executor task specifically, since a planning benchmark would make it look worse than it is. Where did the 4x throughput actually land once the sub-agents were doing real tool calls?","datePublished":"2026-08-11T16:58:12.291Z","url":"https://daily.dev/posts/WcpGEC5lL#c-cZ83AnVQ5","author":{"@type":"Person","name":"kartik-nvjk","url":"https://daily.dev/kartiknvj","image":"https://media.daily.dev/image/upload/s--3gGgsVCw--/f_auto/v1781456774/avatars/avatar_TvTVeiMdkRCqWUDullFmy?_a=BAMAMiWQ0"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1}},{"@type":"Comment","text":"I’d keep this one on git chores and give planning to a model with fewer opinions.","datePublished":"2026-08-12T03:44:28.821Z","url":"https://daily.dev/posts/WcpGEC5lL#c-uV7ZJvsJH","author":{"@type":"Person","name":"Agustin Barrientos","url":"https://daily.dev/agustinbarrientos","image":"https://media.daily.dev/image/upload/s--5ayxQnqn--/f_auto/v1788281802/avatars/avatar_wQYYVe5Tbj0NJ7C7qPoa8?_a=BAMAMicg0"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-fastest-nemotron-yet-embracing-nvidia-nemotron-3-5-lightning-in-kilo-wcpgec5ll#faq","mainEntity":[{"@type":"Question","name":"What is NVIDIA Nemotron 3.5 Lightning and how big is it?","acceptedAnswer":{"@type":"Answer","text":"NVIDIA Nemotron 3.5 Lightning is a customizable open model built as a 30B parameter Mixture-of-Experts architecture with only 3B active parameters, distilled from NVIDIA's frontier Nemotron 3 Ultra model. It supports context windows up to 1M tokens and offers up to 4x higher throughput than some competing models, making it optimized for always-on, always-running agentic workloads. Developers picking a fast open model for agent workloads can compare benchmarks like these on daily.dev."}},{"@type":"Question","name":"What are the Instant and Thinking variants of Nemotron 3.5 Lightning good for?","acceptedAnswer":{"@type":"Answer","text":"The Instant variant handles low-risk, deterministic git operations well, achieving a 100% success rate on tagging plus strong results on undo/reset recovery, commit construction, and remote handling. The Thinking variant achieved a nearly 73% exact completion rate on a git benchmark and scored perfectly (16/16) on deterministic operations like repository inspection, revert, stash, and undo recovery. Teams choosing between model variants for git-heavy agent tasks can track results like these on daily.dev."}},{"@type":"Question","name":"Is Nemotron 3.5 Lightning good for high-level agent planning?","acceptedAnswer":{"@type":"Answer","text":"Not really. It has a highly opinionated, spirited personality that makes it a poor fit for unconstrained, high-level planning, but it excels as a specialized executor when given strict guidelines and bounded instructions, such as within an orchestrator-executor multi-agent setup where it reliably completes clearly defined sub-tasks without drift. Developers designing multi-agent orchestration setups can follow model behavior notes like these on daily.dev."}}]} ```