> For the complete documentation index, see [llms.txt](https://help.intangible.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.intangible.ai/intangible-mcp-agents/nvidia-nemotron.md).

# NVIDIA Nemotron

Intangible works with NVIDIA Nemotron. Nemotron 3.5 Super VL builds 3D scenes in Intangible over MCP when run through Grok Build with NVIDIA's API. Models and setup steps.

**Yes, Intangible works with NVIDIA's open Nemotron models.** Nemotron connects to Intangible's MCP server through an agent app and builds 3D scenes in your Studio tab.

## At a glance

| Setting         | Value                                                                   |
| --------------- | ----------------------------------------------------------------------- |
| Server URL      | `https://mcp.intangible.ai/mcp`                                         |
| Run through     | Grok Build, as a custom model in `~/.grok/config.toml`                  |
| NVIDIA endpoint | `https://integrate.api.nvidia.com/v1`, with `api_backend = "responses"` |
| API key         | `NVIDIA_API_KEY`, from build.nvidia.com                                 |
| Not supported   | Codex. It rejects tool calls from non-OpenAI models                     |

## Models

| Model                                                            | Run through                                                     |
| ---------------------------------------------------------------- | --------------------------------------------------------------- |
| Nemotron 3.5 Super VL (`nvidia/nemotron-3.5-super-vl-120b-a12b`) | Grok Build, NVIDIA API                                          |
| Nemotron 3 Ultra (`nvidia/nemotron-3-ultra-550b-a55b`)           | Should work the same way, depending on your NVIDIA model access |

## Before you start

* An Intangible account. Sign up at [studio.intangible.ai](https://studio.intangible.ai/?utm_source=help.intangible.ai\&utm_medium=docs\&utm_campaign=inline-link\&utm_content=nvidia-nemotron).
* An NVIDIA API key from [build.nvidia.com](https://build.nvidia.com).
* Grok Build installed and signed in. See [xAI Grok](/intangible-mcp-agents/connect-grok.md).

## Set it up

Nemotron is a model. It needs an agent app to call Intangible's tools, and Grok Build works.

1. Get an API key from [build.nvidia.com](https://build.nvidia.com), then set it in the terminal you'll run Grok Build from:

   ```bash
   export NVIDIA_API_KEY="nvapi-..."
   ```

   Add the line to your shell profile (for example `~/.zshrc`) to keep it across sessions.
2. Install Grok Build and connect it to Intangible. See [xAI Grok](/intangible-mcp-agents/connect-grok.md).
3. Add Nemotron as a custom model in `~/.grok/config.toml`:

   ```toml
   [model."nemotron-super-vl"]
   model = "nvidia/nemotron-3.5-super-vl-120b-a12b"
   base_url = "https://integrate.api.nvidia.com/v1"
   name = "Nemotron 3.5 Super VL"
   env_key = "NVIDIA_API_KEY"
   api_backend = "responses"
   ```

   For Nemotron 3 Ultra, add a second entry with the same settings and `model = "nvidia/nemotron-3-ultra-550b-a55b"`.
4. Start an interactive session with `grok -m nemotron-super-vl`, keep one Studio tab open, and ask it to use the intangible MCP server. Then [describe your first scene](/intangible-mcp-agents/intangible-mcp-agents.md#describe-your-first-scene).

## Setup notes

* **Use `api_backend = "responses"`.** With `chat_completions`, Grok Build fails with *"serialization error: invalid type: null, expected u32"*.
* **Run it interactively.** Grok Build signs in to Intangible only inside an interactive session; `grok -p` reports "auth required".
* **Codex can't run Nemotron against Intangible.** It rejects every Intangible tool call from a non-OpenAI model as "unsupported call".

## Related

* [Agents & MCP](/intangible-mcp-agents/intangible-mcp-agents.md)
* [Choosing a model](/intangible-mcp-agents/choosing-a-model.md)
* [xAI Grok](/intangible-mcp-agents/connect-grok.md)

*Last updated 23 October 2026.*


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://help.intangible.ai/intangible-mcp-agents/nvidia-nemotron.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.