Documentatation – Eagle Servers

Documentation

Everything you need to build with eagleservers

Quick Start

Get up and running with EagleServers in under 5 minutes.

i. Install the CLI

npm install -g @eagleservers/cli

# or

curl -sSL https://get.eagleservers.com | bash

ii. Authenticate

eagleservers auth login

# Enter your API key when prompted

iii. Deploy Your First Instance

eagleservers deploy --name my-first-app \

  --type gpu-rtx4090 \

  --region us-west-1 \

  --image ubuntu:22.04

Pro Tip: Use eagleservers deploy --help to see all available options.

Installation

Choose your preferred installation method:

Node.js — Python — Docker

// Install SDK

npm install @eagleservers/sdk

// Usage

const EagleServers = require('@eagleservers/sdk');

const client = new EagleServers({

  apiKey: 'your-api-key'

});

// Create instance

const instance = await client.instances.create({

  name: 'gpu-worker',

  type: 'gpu-rtx4090',

  region: 'us-west-1'

});

Authentication

All API requests require authentication using your API key.

i. Getting Your API Key

Sign in to your dashboard
Navigate to Settings → API Keys
Click "Generate New Key"
Store your key securely

ii. Using Your API Key

# Via environment variable

export EAGLESERVERS_API_KEY="your-api-key"

# Via CLI flag

eagleservers --api-key="your-api-key" instances list

# Via config file

echo "api_key: your-api-key" > ~/.eagleservers/config.yaml

GPU Computing Guide

Optimize your GPU workloads for maximum performance.

Available GPU Types

GPU Model	VRAM	CUDA Cores	Best For	Price/hr
RTX 4090	24GB	16,384	Large models, training	$1.79
A100	80GB	6,912	Enterprise ML	$2.79
T4	16GB	2,560	Inference	$0.55
V100	32GB	5,120	Research	$1.29

CUDA Setup

import torch

import eagleservers

# Check CUDA availability

print(f"CUDA Available: {torch.cuda.is_available()}")

print(f"GPU Count: {torch.cuda.device_count()}")

# Initialize model on GPU

model = YourModel().cuda()

# Distributed training

if torch.cuda.device_count() > 1:

  model = nn.DataParallel(model)