> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nexa.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Quickstart
> Get started with Nexa AI SDK for Android in minutes.
## **Tutorial Video**
## **Installation**
Nexa AI Android SDK is available from Maven Central, the latest version is [0.0.24](https://repo1.maven.org/maven2/ai/nexa/core/0.0.24/). You can also find it on [Maven Repository](https://mvnrepository.com/artifact/ai.nexa/core) which updates a bit slower than Maven Central.
### **Add Dependency**
Add the following to your `app/build.gradle.kts` to import NexaSDK:
```kotlin theme={"dark"}
android{
packagingOptions {
jniLibs.useLegacyPackaging = true
}
}
dependencies {
implementation("ai.nexa:core:0.0.24")
}
```
The default repository is mavenCentral(). You can also add github's raw repository in the settings.gradle.kts:
```kotlin theme={"dark"}
dependencyResolutionManagement{
repositories {
maven {
url = uri("https://raw.githubusercontent.com/NexaAI/core/main")
}
}
}
```
## **Run Your First Model**
Initialize the NexaSDK in your Android application:
```kotlin theme={"dark"}
NexaSdk.getInstance().init(this)
```
Create a VLM wrapper and load a model for NPU inference. You can download [OmniNeural-4B model](https://huggingface.co/NexaAI/OmniNeural-4B-mobile) from huggingface and place it in the folder that your app has permission to access, for example: `/data/data/${your-app-packagename}/files/models/OmniNeural-4B`
```kotlin theme={"dark"}
VlmWrapper.builder()
.vlmCreateInput(
VlmCreateInput(
model_name = "omni-neural",
model_path = ,
config = ModelConfig(
max_tokens = 2048,
enable_thinking = false
),
plugin_id = "npu"
)
)
.build()
.onSuccess { vlmWrapper = it }
```
Your app should have permission to access the model folder. For example, in the demo, the model\_path is set to `/data/data/com.nexa.demo/files/models/OmniNeural-4B/files-1-1.nexa`
Use the loaded model to generate responses:
```kotlin theme={"dark"}
vlmWrapper.generateStreamFlow("Who are you?", GenerationConfig()).collect { result ->
println(result)
}
```
## **Supported Models**
We've curated collections of compatible models for you to explore:
Optimized models for Qualcomm NPU
Any GGUF format LLM and VLM models
## **Device Compatibility**
* Qualcomm Snapdragon 8 Elite
* Qualcomm Snapdragon 8 Elite Gen 5
* Qualcomm Adreno GPU
* ARM64-v8a
## **Next Steps**
Explore detailed documentation for all model types and learn their usage.