NVIDIA is a US semiconductor and computing company headquartered in Santa Clara, California, and listed on Nasdaq. It is best known for GPUs and the CUDA platform used to train AI models, and it also publishes models such as the Nemotron family.
- Founded
- 1993
- Type
- Public company
- Founders
- Jensen Huang, Chris Malachowsky and Curtis Priem
- AI lab
- Inference provider
- 0 hosted endpoints
Catalogue
All models by NVIDIA
32 models · 26–32
| Select models to compare | Model | Released | Context | Input /M | Output /M | Parameters | Type | I/O | Openness |
|---|---|---|---|---|---|---|---|---|---|
 Nemotron Nano 9B v2NVIDIA | August 18, 2025 | 131K | — | — | 9B | Text | text → text | Open weights | |
| Llama 3.3 Nemotron Super 49B v1.5NVIDIA | July 25, 2025 | 131K | — | — | 49B | Text | text → text | Open weights | |
| Mistral NemotronNVIDIA | June 11, 2025 | 128K | — | — | — | Text | text → text | Closed | |
| Llama 3.1 Nemotron 70B InstructNVIDIA | April 15, 2025 | 128K | — | — | 70B | Text | text → text | Open weights | |
| Llama 3.1 Nemotron Ultra 253BNVIDIA | April 7, 2025 | 128K | — | — | 253B | Text | text → text | Open weights | |
| Llama 3.3 Nemotron Super 49B v1NVIDIA | March 18, 2025 | 131K | — | — | 49B | Text | text → text | Open weights | |
| Nemotron Mini 4B InstructNVIDIA | August 21, 2024 | 128K | — | — | 4B | Text | text → text | Open weights |