# Nemotron 3 Super

20 Hacker News commenters recommended Nemotron 3 Super and 3 criticized it, November 2025 to November 2026. It was recommended in 7 of 13 months. It meets the Hacker Approved badge rule.

**Verdict:** Open-weight Nvidia model that people run locally and praise for fast inference and open training data. Critics say it trails Qwen models, especially at coding.

## Why people recommend it

- Fast inference and high throughput (4 comments): "It offers ~same capabilities as gpt-oss-120b but with faster inference" ([source](https://news.ycombinator.com/item?id=47346647)) "Nvidia's nemotron recent 120B model also has some really crazy tps" ([source](https://news.ycombinator.com/item?id=50148549)) "Nemotron-3-Super: 14070 tok/s prefill, ~194.5 tok/s generation." ([source](https://news.ycombinator.com/item?id=47620792))
- Open training data and recipes (3 comments): "I do really like that they have been transparent in releasing most of its training data." ([source](https://news.ycombinator.com/item?id=47369087)) "probably most important: training datasets and training recipe available" ([source](https://news.ycombinator.com/item?id=47349118)) "Nvidia provides almost everything you need to replicate it from scratch" ([source](https://news.ycombinator.com/item?id=49421890))
- Runs well on local hardware (2 comments): "running nearly everything I do locally" ([source](https://news.ycombinator.com/item?id=47368524)) "we can task our nemotron-3-super model to run over night when our desktops are idle" ([source](https://news.ycombinator.com/item?id=47432312))
- Holds up in agentic, long-context work (2 comments): "this thing has an ability to hold attention through complicated (often noisy) agentic environments" ([source](https://news.ycombinator.com/item?id=47620217)) "Nemotron super 3 110B works well for 1M context long vibecoding sessions" ([source](https://news.ycombinator.com/item?id=48546316))

## What people warn about

- Weaker at coding than Qwen (2 comments): "Nemotron Super 120b is great at a lot of stuff but it also seems to be not as good at coding as Qwen." ([source](https://news.ycombinator.com/item?id=48544883)) "does poorly on agentic coding that I've given it" ([source](https://news.ycombinator.com/item?id=48090008))
- Trails newer Qwen models overall (2 comments): "well behind of Qwen 3.8 flash next on pretty much everything except generation speed" ([source](https://news.ycombinator.com/item?id=50104536)) "this 3.5 Super is still below Qwen 3.8 next as they state themsleves" ([source](https://news.ycombinator.com/item?id=50105927))

## Alternatives hackers recommend

- [Qwen 3.8 27B](https://hackerapproved.com/p/qwen-3-8-27b.md): 1173 recommend
- [Z.ai GLM](https://hackerapproved.com/p/glm-5-2.md): 1043 recommend
- [Kimi](https://hackerapproved.com/p/kimi-k3.md): 931 recommend
- [Gemma 4](https://hackerapproved.com/p/gemma-4.md): 713 recommend
- [Qwen](https://hackerapproved.com/p/qwen.md): 474 recommend

## Cite as

Hacker Approved, "Nemotron 3 Super", https://hackerapproved.com/p/nemotron-3-super/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).