NVIDIA's MoE with 12B active parameters for coding, reasoning, tool use and document understanding. Runs on a single H200 or a pair of A100/H100.
| GPQA Diamond | 89.5 | |
|---|---|---|
| LiveCodeBench v6 | 88.2 | |
| SWE-bench Verified | 71.1 |
Which GPU does Nemotron 3.5 Super 120B-A12B need?
The entry configuration is 1x H200, which delivers around 35 tokens per second. For more parallel requests and headroom on speed, go with 2x H200.
How much does Nemotron 3.5 Super 120B-A12B cost?
An hour of Nemotron 3.5 Super 120B-A12B runs from $3.86 to $8.28 depending on the GPU configuration. Billing is hourly: you pay for server time, not for tokens.
How do I call Nemotron 3.5 Super 120B-A12B from my code?
Once deployed you get an OpenAI-compatible endpoint: point any OpenAI client at the base_url you receive and pass nemotron-3.5-super:120b as the model name. Code written against the OpenAI API needs no changes.
Does the data stay private?
Yes. Nemotron 3.5 Super 120B-A12B runs on a dedicated GPU server inside a private environment: requests and responses never reach the model vendor and are not used for training.
What context window and license does Nemotron 3.5 Super 120B-A12B have?
Nemotron 3.5 Super 120B-A12B holds 32K tokens of context per request and ships under the OpenMDW-1.1 license, so you can use it commercially on your own hosting terms.