1. 
[Home](/)
2. [RTX](/rtx)

In-Game Inferencing

# NVIDIA In-Game Inferencing SDK

The NVIDIA In-Game Inferencing (NVIGI) SDK offers a streamlined and high performance path to integrate locally run AI models into games and applications via in-process (C++) execution and CUDA in Graphics. NVIGI supports all major inference backends, across different hardware accelerators (GPU and CPU), so developers can take advantage of the full range of available system resources on a user’s PC.

NOTE: both the x64 and the ARM64 packs will run on the [NVIDIA RTX Spark](https://www.nvidia.com/en-us/products/rtx-spark/) platform. The ARM64 pack will work for applications natively compiled for the ARM64 architecture, while the x64 pack will work for applications compiled for the x64 architecture and then run in x64 emulation mode.

[Download x64](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.7/nvigi_pack_1_7_0_x64.7z)[Download ARM64](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.7/nvigi_pack_1_7_0_arm64.7z)

 ![NVIDIA In-Game Inferencing SDK stack diagram](https://developer.download.nvidia.com/images/rtx/NvIGI-SDK-diagram.png)

* * *

## Benefits of NVIDIA In-Game Inferencing SDK

### Scalable Model Support

NVIDIA In-Game Inferencing supports speech, audio, visual and language models for in-process (C++) execution.

### Simplified Deployment

Load and run models built for TensorRT, ONNX Runtime, Llama.cpp or a custom executor that works across GPU and CPU via a unified inference API.

### 
### Accelerated Performance  

Ensures that CUDA in Graphics (CiG) is properly set up to optimize graphics and AI execution within game pipelines, allowing them to happen simultaneously at low latency.

* * *

## NVIDIA In-Game Inferencing Plugins &amp; Model Support

The NVIDIA In-Game Inferencing SDK provides plugin support for a number of different language and speech models including:

| NVIGI Plugin | Supported Inference Hardware | Supported Models |
| --- | --- | --- |
| Speech - ASR Local TRT | CUDA-Enabled GPU | 
- [Riva ASR](/downloads/ace/speech-model/riva_asr_pack_1.1.zip) V2

 |
| Speech - TTS Local TRT | CUDA-Enabled GPU | 
- [Riva TTS](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/nvigisdk/models/riva-tts-a2flow)
- [Chatterbox Turbo 350M TTS](/downloads/ace/speech-model/chatterbox_turbo.zip)
- [Chatterbox Multilingual 500M TTS](/downloads/ace/speech-model/chatterbox_multilingual.zip)

 |
| 

Language - GPT Local GGML | 

CUDA-Enabled GPU or CPU | 
- [Nemotron 3 Nano 4B](/downloads/assets/ace/intelligence/nemotron_4b_igi.zip)

- [Nemotron 2 Nano 9B](/downloads/assets/ace/model_zip/nvidia_nvidia-nemotron-nano-9b-v2-q4_k_m.gguf.zip)

- [Qwen3-0.6B](https://developer.nvidia.com/downloads/assets/ace/model_zip/qwen_qwen3-0.6b-q4_k_m.gguf.zip)

- [Qwen3,5-4B](/downloads/assets/ace/model_zip/qwen3-4b-q4_k_m.gguf.zip)

- [Qwen3.5-8B](/downloads/assets/ace/model_zip/qwen3-8b-q4_k_m.gguf.zip)

 |
| RAG - Embed Local GGML | CUDA-Enabled GPU or CPU | 
- [[E5 Large Unsupervised](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/nvigisdk/models/e5-large-unsupervised)](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/nvigisdk/models/e5-large-unsupervised)

 |

Developers can also use the open source NVIGI SDK to build their own plugins for models not listed above.

* * *

## Related Products  

 ![NVIDIA RTX AI toolkit](https://developer.download.nvidia.com/images/rtx/aim-related-toolkit-1920x1080.jpg)

### NVIDIA ACE  

NVIDIA ACE is a suite of digital human technologies that bring game characters and digital assistants to life with generative AI.

[Learn More About ACE](https://developer.nvidia.com/ace)

 ![NVIDIA ACE sparks life into virtual characters with generative AI](https://developer.download.nvidia.com/images/rtx/covertprotocol-shots-ari.jpg)

### NVIDIA ACE Unreal Engine 5 Reference

Plugins and samples for Unreal Engine developers looking to bring their MetaHumans to life with generative AI on RTX PCs.

[Learn More About ACE Unreal Engine 5 Reference](https://developer.nvidia.com/ace/get-started#plugins)

* * *

## Resources  

### 
### How to Successfully Integrate NVIDIA ACE for Autonomous Game Characters

This guide explores how NVIGI integrates with ACE to enable seamless AI inference in game development. It covers NVIGI’s architecture, key features, and how to get started creating autonomous characters with NVIDIA ACE on-device models.

[Read Blog](https://developer.nvidia.com/blog/bring-nvidia-ace-ai-characters-to-games-with-the-new-in-game-inference-sdk/)

### 
### FAQ

Find answers to frequently asked questions on NVIDIA In-Game Inferencing FAQ.

[See FAQ](https://forums.developer.nvidia.com/t/ace-igi-faq/323518)

### 
### Additional Advanced Use Components

While the downloaded binary pack will suffice for most applications, NVIGI also includes some additional components for advanced development cases. The most common of these cases is NVIGI developers creating their own plugins.

[NVIGI Repositories](https://github.com/NVIDIA-RTX/NVIGI)

### Unreal Engine 5 Sample

A minimalistic code sample that shows how to integrate AI features (GPT/LLM) into an Unreal Engine 5 application.

[Access Sample](https://github.com/NVIDIA-RTX/NVIGI-UEPlugin)

* * *

## Archive

- [NVIGI 1.6.1](https://developer.nvidia.com/downloads/rtx/In-Game-Inference-SDK/1.6.1/nvigi_pack_1_6_1.7z)
- [NVIGI 1.6](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.6/nvigi_pack_1_6_0.7z)
- [NVIGI 1.5 GDC](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.5-gdc/nvigi_pack_1_5_0_gdc.7z)
- [NVIGI 1.5](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.5/nvigi_pack_1_5_0.7z)
- [NVIGI 1.4](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.4/nvigi_pack_1_4_0.7z)
- [NVIGI 1.3](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.3/nvigi_pack_1_3_0.7z)
- [NVIGI 1.2](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.2/nvigi_pack_1_2_0.7z)
- [NVIGI 1.0](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/nvigi_pack_ga_1_0_0.zip)
- [NVIGI 1.1.0](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.1.0/nvigi_pack_1_1_0.zip)
- [NVIGI 1.1.1](https://developer.nvidia.com/downloads/rtx/In-Game-Inference-SDK/1.1.1/nvigi_pack_1_1_1.zip)

Get started with the NVIDIA In-Game Inferencing SDK.

[Download x64](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.7/nvigi_pack_1_7_0_x64.7z)[Download ARM64](https://developer.nvidia.com/downloads/rtx/in-game-inference-sdk/1.7/nvigi_pack_1_7_0_arm64.7z)


