Model Library
NVIDIAReleased March 11, 2026NVIDIA Nemotron 3 Super
NVIDIA · Released March 11, 2026
Intelligence score
25
At a glance
- Intelligence
- 25
- Speed
- 363.7tokens/sec
- Latency
- 1.37stime to first token
- Input price
- $0.25per 1M tokens
- Output price
- $0.78per 1M tokens
- Context window
- 3tokens
- Parameters
- 120.6billions
Summary
Comparison summary
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) scores 25 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight models of similar size (median: 9). NVIDIA Nemotron 3 Super 120B A12B (Reasoning) generates output at 363.7 tokens per second (based on the median across providers serving the model), which is well above average compared to other open weight models of similar size (median: 88.2 t/s). NVIDIA Nemotron 3 Super 120B A12B (Reasoning) costs $0.25 per 1M input tokens (better than average, median: $0.40) and $0.78 per 1M output tokens (better than average, median: $0.84), based on the median across providers serving the model.
Specifications
Technical specifications
| Model name | NVIDIA Nemotron 3 Super |
|---|---|
| Provider | NVIDIA |
| Release date | March 11, 2026 |
| Intelligence score | 25 |
| Speed | 363.7 tokens/sec |
| Latency (TTFT) | 1.37s |
| Input price | $0.25 / 1M tokens |
| Output price | $0.78 / 1M tokens |
| Context window | 3 tokens |
| Parameters | 120.6B |
| Reasoning | Yes |
| Input modalities | text |
| Output modalities | — |
| Open weights | Yes |
| License | Nvidia Nemotron Open Model License |
| API availability | — |
| Source | Model Intelligence legacy profile: nvidia-nemotron-3-super-120b-a12b |
Pricing
Token prices
Price per 1M tokens
NVIDIA Nemotron 3 Super, as listed in the catalog record
- Input$0.25
- Output$0.78
| Token type | Price per 1M |
|---|---|
| Input | $0.25 |
| Output | $0.78 |
Model Intelligence legacy profile: nvidia-nemotron-3-super-120b-a12b
FAQ · 21
Frequently asked questions
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) was released on March 11, 2026.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) was created by NVIDIA.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) scores 25 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight models of similar size (median: 9).
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) generates output at 363.7 tokens per second (based on the median across providers serving the model), which is well above average compared to other open weight models of similar size (median: 88.2 t/s).
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) has a time to first token (TTFT) of 1.37s (based on the median across providers serving the model), which is better than average compared to other open weight models of similar size (median: 1.52s).
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) costs $0.25 per 1M input tokens (better than average, median: $0.40) and $0.78 per 1M output tokens (better than average, median: $0.84), based on the median across providers serving the model.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) costs $0.25 per 1M input tokens and $0.78 per 1M output tokens (based on the median across providers serving the model). For a blended rate (7:2:1 cache hit/input/output ratio), this is $0.23 per 1M tokens. Pricing may vary by provider.
When evaluated on the Intelligence Index, NVIDIA Nemotron 3 Super 120B A12B (Reasoning) generated 110M output tokens, which is at the higher end compared to other open weight models of similar size (median: 59M).
Yes, NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) supports text input.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) supports text output.
No, NVIDIA Nemotron 3 Super 120B A12B (Reasoning) does not support image input. It can only process text.
No, NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is not multimodal. It only supports text input.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) has a context window of 1.0M tokens. This determines how much text and conversation history the model can process in a single request.
Yes, NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is open weights. The model weights are publicly available and can be downloaded for self-hosting.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) has 120.6 billion parameters (12.7 billion active).
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is a Mixture of Experts (MoE) model with 120.6 billion total parameters, but only 12.7 billion active parameters are used during inference.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is released under the Nvidia Nemotron Open Model License license. This license allows commercial use.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) achieves a score of 25 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.
Yes, NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is available via API through 4 providers.
NVIDIA Nemotron 3 Super 120B A12B (Reasoning) is available through 4 API providers.