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

# Quickstart

> Create an account, deploy a model, and run your first inference request in minutes

This guide walks you from a new account to a working API call. Expect about 10–15 minutes for account setup and model deployment.

## Prerequisites

* A [Google](https://accounts.google.com) or [GitHub](https://github.com/login) account for sign-up
* A Hugging Face account with a model in [Safetensors](https://huggingface.co/docs/safetensors) format, or use a public model from the Hub

## Step 1: Create your account

1. Go to [synapsai.cloud/login](https://synapsai.cloud/login).
2. Sign in with Google or GitHub.

<Tip>
  Planning to collaborate with teammates? Create a [team](/guides/team) and invite members before deploying shared models.
</Tip>

## Step 2: Review your credits

SynapsAI Cloud includes starter credits for new accounts. Check your balance on the [Billing](https://platform.synapsai.cloud/billing) page before deploying.

See [Billing](/manage/billing) for purchasing additional credits and enabling auto-pay.

## Step 3: Understand core concepts

Before deploying, read [Core concepts](/core-concepts) — especially **readiness levels** and **pricing**. Estimated costs are shown in the deployment UI before you confirm.

## Step 4: Deploy a model

<Card title="Launch a model" icon="rocket" href="https://platform.synapsai.cloud/launch-model">
  Open the deployment wizard in the dashboard.
</Card>

Follow [Deploy a model](/guides/deploy-model) for Hugging Face requirements, readiness levels, precision, and scaling options.

If you need to prepare a repository first, see [Launch a custom model](/guides/hf-model-setup) or [Convert models to Safetensors](/guides/convert-safetensors).

Wait until deployment finishes — the model status should be **Sleeping** or **Ready**, not **Building**. You can send inference requests in either state. When **Sleeping**, the platform automatically loads an instance on your request.

## Step 5: Create an API key

1. Open [API keys](https://platform.synapsai.cloud/apikeys).
2. Click **Create Secret Key**.
3. Name the key and restrict it to your newly deployed model.
4. Click **Create Key** and copy the value — it is shown only once.

Store the key securely. See [API keys](/manage/api-keys) for rotation and permission guidance.

```bash theme={null}
export SYNAPSAI_API_KEY="your-api-key"
```

## Step 6: Run inference

Install the Python SDK:

```bash theme={null}
pip install --upgrade synapsai-python
```

Send a chat completion request. Replace `your-model-id` with the ID from the [Models](https://platform.synapsai.cloud/models) page:

```python theme={null}
from synapsai import SynapsAI

client = SynapsAI()

response = client.chat.completions.create(
    model="your-model-id",
    messages=[{"role": "user", "content": "Hello! What can you help me with?"}],
)

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

Or use cURL:

```bash theme={null}
curl https://api.synapsai.cloud/v1/chat/completions \
  -H "Authorization: Bearer $SYNAPSAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "your-model-id",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

For streaming, tool calling, and other pipelines, see [Inference quickstart](/guides/inference-quickstart) and [Examples](/examples/introduction).

## Next steps

<Columns cols={2}>
  <Card title="Migrate from OpenAI" icon="arrow-right-arrow-left" href="/guides/migrate-from-openai">
    Use OpenAI-compatible clients with SynapsAI.
  </Card>

  <Card title="API reference" icon="code" href="/api-reference/introduction">
    Authentication, errors, and endpoint details.
  </Card>

  <Card title="Optimize costs" icon="dollar-sign" href="/guides/optimizing-costs">
    Tune readiness, quantization, and context length.
  </Card>

  <Card title="Integrations" icon="plug" href="/guides/integrations">
    LangChain, LlamaIndex, FAISS, and more.
  </Card>
</Columns>