VPS for Algo Trading India 2026 — Specs, Mumbai Latency, and Setup
Algo trading in India has exploded — Zerodha's Kite Connect, Upstox API, Angel One SmartAPI, Dhan API, and Fyers API have made systematic trading accessible to retail traders and developers. The most common question after setting up a strategy: should I run it on my laptop or on a VPS? And if VPS — which specs and which location?
This guide covers what you actually need for an Indian algo trading VPS — minimum specs, Mumbai vs other cities, the correct Python/Node.js setup, and why ₹849/mo is often all a solo trader needs.
Why Mumbai VPS is the only right answer for Indian algo trading
All major Indian brokers (Zerodha, Upstox, Angel One, Dhan, Fyers, Shoonya) have their API infrastructure in Mumbai — adjacent to NSE's data centre in BKC and BSE's infrastructure in Fort. Running your bot from:
- Mumbai VPS: 5–20ms RTT to broker API endpoints
- Bangalore VPS: 20–40ms (adds 15–20ms vs Mumbai)
- Singapore VPS: 80–110ms (significant lag for tick-sensitive strategies)
- Your home internet (Delhi/Chennai/Hyderabad): 30–80ms + ISP instability
For most retail algo trading strategies (not HFT co-location), the 15–20ms difference between Mumbai and Bangalore does not matter for execution. But VPS uptime and ISP stability matter a lot — which is why any VPS beats a home connection for 24/7 bot operation.
Minimum VPS specs for Indian algo trading
| Setup | Specs needed | HostStack plan | Price |
|---|---|---|---|
| 1 strategy, 1 broker API, Python/Node | 1 vCPU / 2 GB / 40 GB | X1 Shared VPS | ₹849/mo |
| 3–5 strategies, pandas/numpy, live data | 2 vCPU / 4 GB / 80 GB | X2 Shared VPS | ₹1,199/mo |
| 10+ strategies + DB + dashboard | 4 vCPU / 8 GB / 160 GB | X3 Shared VPS | ₹1,999/mo |
| Production trading desk, low latency critical | 4 vCPU dedicated / 8 GB | X3 VDS (dedicated) | ₹3,999/mo |
All prices ex-GST, Mumbai Ryzen KVM. VDS = virtual dedicated server with dedicated CPU cores — better for latency-sensitive workloads.
Python algo trading setup on Ubuntu VPS
Install the trading stack
# Install Python 3.12 + pip
sudo apt update && sudo apt install -y python3 python3-pip python3-venv
# Create project environment
python3 -m venv ~/trading-env
source ~/trading-env/bin/activate
# Install common trading libraries
pip install kiteconnect # Zerodha Kite Connect SDK
pip install smartapi-python # Angel One SmartAPI
pip install pandas numpy # Data manipulation
pip install python-dotenv # Environment variable management
pip install schedule # Job scheduling
pip install requests websocket-client
Run your bot with PM2 (keeps it alive 24/7)
npm install -g pm2
pm2 start "source ~/trading-env/bin/activate && python bot.py" \
--name trading-bot \
--interpreter bash
pm2 startup # enable restart on server reboot
pm2 save
pm2 logs trading-bot # view real-time logs
Systemd alternative (no npm needed)
# /etc/systemd/system/trading-bot.service
[Unit]
Description=Algo Trading Bot
After=network.target
[Service]
User=ubuntu
WorkingDirectory=/home/ubuntu/trading
ExecStart=/home/ubuntu/trading-env/bin/python bot.py
Restart=always
RestartSec=5
[Install]
WantedBy=multi-user.target
# Enable: sudo systemctl enable trading-bot && sudo systemctl start trading-bot
Critical setup for trading bots on VPS
- Timezone: Set your VPS timezone to IST:
sudo timedatectl set-timezone Asia/Kolkata. Your cron jobs and scheduled tasks need to run at IST market hours. - Auto-login tokens: Zerodha Kite requires a new access token daily. Automate this with a cron job at 8:00 AM IST that generates the token before market open (9:15 AM).
- Alert on crash: Configure PM2 or systemd to notify you on Telegram or email if the bot crashes.
pm2 install pm2-telegramor a simple bash script that hits a Telegram bot API. - Never run as root: Create a dedicated user:
sudo adduser trading. Run the bot as that user. - Backups: Schedule daily rsync of your strategy code and trade log to an off-VPS location (even a personal Google Drive via rclone).
VDS vs shared VPS for trading — which to pick?
For most retail traders: Ryzen Shared VPS is fine. The CPU sharing on shared VPS adds milliseconds of jitter — negligible for strategies that operate on minutes/hours timescales.
Consider Ryzen VDS (dedicated cores) if:
- You run tick-by-tick strategies where CPU scheduling jitter matters
- You process large historical datasets or run backtests on the same VPS
- You run 10+ concurrent WebSocket connections with low-latency event handling
DDoS protection — why it matters for trading VPS
On budget day, election results, and major earnings announcements — Indian financial forums and social media channels see coordinated DDoS attacks on broker infrastructure and trading-adjacent services. If your VPS hosts a trading dashboard or API relay, it can become a target.
HostStack includes DDoS-aware network protection on all plans — your Mumbai VPS has edge filtering before traffic reaches your server. This is not always true on the cheapest Indian VPS providers. Check before you buy.