AI model · NVIDIA
Nemotron 3 Ultra (free)
Model ID: nvidia/nemotron-3-ultra-550b-a55b:free
- Input
- Free per 1M tokens
- Output
- Free per 1M tokens
- Context
- 1M tokens at once
- Max output
- 66K tokens per reply
- Released
- 4 Jun 2026 NVIDIA
What it can do
- YesReasoning
- YesTool calling
- NoImage input
- YesOpen weights
- YesFree to use
Takes text; gives text.
How its listing describes it
NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...
Sources
- NVIDIA: Nemotron 3 Ultra (free) on OpenRouter — read 2 Oct 2026