# Nim

> Invoke various LLMs (GLM-5, Kimi-k2.5, Llama 3.1, etc.) via NVIDIA NIM API to save main agent tokens and leverage specialized model capabilities.

- Skill: `d-wwei/nim` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add d-wwei/nim`
- Raw SKILL.md: https://api.skillmd.com/api/skills/d-wwei/nim/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: d-wwei (https://skillmd.com/u/d-wwei)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/d-wwei/nim

---


# NVIDIA NIM Skill for OpenClaw

This skill allows OpenClaw to delegate tasks to external models hosted on the NVIDIA NIM platform.

## Setup

1. **Get API Key**: Register at [build.nvidia.com](https://build.nvidia.com) and get your `nvapi-...` key.
2. **Set Environment Variable**:
   ```bash
   export NVIDIA_API_KEY="your_api_key_here"
   ```

## Usage

### 查看可用模型
```bash
python3 scripts/nim_call.py list
```

### 调用模型
```bash
python3 scripts/nim_call.py <model_alias> "<prompt>"
```

## 支持的常用别名
(运行 `list` 命令查看完整实时列表)
- `glm5`, `kimi`, `r1-qwen-32b`, `llama33`, `phi4`, `qwq`

## Integration with CLAUDE.md
Add this to your project's `CLAUDE.md`:
```markdown
- /nim <model> <prompt>: Call NVIDIA NIM models.
```