Nemotron 3 Super
Nemotron 3 Super is a large language model made by NVIDIA, added to the catalog on 11 Mar 2026. It has a context window of 262,144 tokens and costs $0.075 per 1M input tokens and $0.40 per 1M output tokens. It accepts text and produces text. Typical uses: analyzing long documents, generating structured responses and working with tool-integrated workflows.
Price checked ยท source: OpenRouter
What Nemotron 3 Super is
NVIDIA Nemotron 3 Super is a text-in, text-out model from NVIDIA. Its catalog description identifies it as an open hybrid mixture-of-experts model with a hybrid Mamba-Transformer design. The description refers to a 120-billion-parameter model with 12 billion parameters activated. It is presented for complex multi-agent applications, while its listed input and output modalities are text.
The model accepts up to 262,144 tokens of context and can produce up to 16,384 output tokens. Its context capacity can suit workflows that provide lengthy documents, detailed instructions, or substantial conversation history in one request. The output limit can accommodate extended responses or generated material. The catalog also lists reasoning support, making analysis-oriented prompts a possible fit alongside general text generation.
Nemotron 3 Super supports tools, structured outputs, and response-format controls. These capabilities can suit workflows that use tool integrations or need responses in a defined structure for another application. The catalog does not specify particular tools, schemas, or output formats. Since the listed modalities are text only, the model is described here for text-based tasks rather than image or audio input and output. Potential uses include examining long text, drafting structured responses, and supporting multi-step or multi-agent workflows that use reasoning and tools.
Key features
- Accepts and produces text
- Supports a context length of 262,144 tokens
- Supports up to 16,384 output tokens
- Includes reasoning and tool support
- Supports structured outputs and response formats
Typical uses
- Analyzing long documents
- Generating structured responses
- Working with tool-integrated workflows
- Supporting multi-step text tasks
Nemotron 3 Super pricing
US dollars, list price through OpenRouter.
| Item | Standard |
|---|---|
| Input, per 1M tokens | $0.075 |
| Output, per 1M tokens | $0.40 |
No price change since it was added on 11 Mar 2026.
| What a task costs | Cost |
|---|---|
| Answer a short question500 in / 300 out | $0.0002 |
| Summarise a 50-page PDF25,000 in / 800 out | $0.0022 |
| Review a 2,000-line source file30,000 in / 2,000 out | $0.0031 |
| Read a 300-page contract set150,000 in / 1,500 out | $0.012 |
Nemotron 3 Super specifications
| Model id | nvidia/nemotron-3-super-120b-a12b |
|---|---|
| Version | nemotron-3-super-120b-a12b-20230311 |
| Context window | 262,144 tokensAbout 196,608 English words |
| Max output | 16,384 tokens |
| Knowledge cutoff | Not published |
| Added to catalog | 11 Mar 2026 |
| Input types | text |
|---|---|
| Output types | text |
| Tool calling | Yes |
| Structured outputs | Yes |
| Reasoning effort control | Yes |
| Content moderation | Not applied by provider |
Sample outputs
The same prompts are given to every model, and the unedited output is shown with the date it was produced. Upend does not publish leaderboard scores it did not run.
No output has been recorded for Nemotron 3 Super yet.
Compare Nemotron 3 Super
Side by side with other models: prices, context window, what each accepts and supports, and the same four tasks priced on both.