> ## Documentation Index
> Fetch the complete documentation index at: https://docs.apus.network/llms.txt
> Use this file to discover all available pages before exploring further.

# APUS AI Inference API 

> A concise guide to building with the APUS AI Inference API on the AO Network

## Introduction

**APUS AI Inference** is an OpenAI-compatible API layer designed for **Deterministic**, **Confidential**, and **Verifiable** AI computation.

It runs on the **Arweave AO decentralized compute network**, combining trusted execution with reproducible model outputs.

***

## **Core Features**

* **Deterministic** — Reproducible inference results across trusted nodes
* **Confidential** — Secure model execution with data privacy guarantees
* **Verifiable** — Transparent computation trace and on-chain attestations
* **OpenAI-Compatible** — Works directly with the official openai SDK

**Install the SDK**

```bash theme={null}
pip install openai
```

**Initialize the Client**

```python theme={null}
from openai import OpenAI

client = OpenAI(
    api_key="",  # Not required during testing
    base_url="https://hb.apus.network/~inference@1.0",
)

MODEL = "google/gemma-3-27b-it"
```

**Single-turn Chat Example**

```python theme={null}
resp = client.chat.completions.create(
    model=MODEL,
    messages=[{"role": "user", "content": "What is 2 + 2?"}],
)

print(resp.choices[0].message.content)
```

**Multi-turn Conversation Example**

```python theme={null}
messages = [
    {"role": "system", "content": "You are a math assistant."},
    {"role": "user", "content": "What is 10 * 10?"},
]

resp = client.chat.completions.create(model=MODEL, messages=messages)
print("Assistant:", resp.choices[0].message.content)

# Continue the conversation
messages += [
    {"role": "assistant", "content": resp.choices[0].message.content},
    {"role": "user", "content": "And what is 100 / 5?"},
]

resp2 = client.chat.completions.create(model=MODEL, messages=messages)
print("Assistant:", resp2.choices[0].message.content)
```

## **Notes**

* No API key required during the **test phase**
* Replace base\_url with your custom endpoint
* Fully compatible with openai.chat.completions.create()
* Designed for **Deterministic**, **Confidential**, and **Verifiable** AI inference