Text Generation
Transformers
Safetensors
PyTorch
nemotron_h
nvidia
nemotron-3
latent-moe
mtp
conversational
custom_code
Eval Results

I tested Nemotron 3 Super in an Agentic Workflow [Video + Report]

#17
by zacksiri - opened

Based on my testing, these models do not perform well at all. They fail even the basic tests. I'm pasting my report here:
https://www.youtube.com/watch?v=5CES0NUNK4I

Link to the full report is here:
https://upmaru.com/llm-tests/simple-tama-agentic-workflow-q1-2026/nemotron-3-super-120b

zacksiri changed discussion title from I tested Nemotron 3 Super in an Agentic Workflow to I tested Nemotron 3 Super in an Agentic Workflow [Video + Report]

Sign up or log in to comment