How to Build an AI Astrology Chatbot in 30 Minutes

About the author: Brett Calloway is a Developer Advocate and AI Integration Specialist with 12 years of experience building APIs and developer tooling, including three years focused on AI-native infrastructure for spiritual and wellness applications. He writes on building context-rich AI agents, drawing on a developer relations background, with coverage spanning API integration patterns, AI agent architecture, and rapid prototyping with astrology, tarot, and numerology data.

Why most AI astrology chatbots are fake

Open any popular AI assistant and ask for your full birth chart. It will misplace Saturn by an entire sign, invent aspects that do not exist, and produce a Vedic kundli with no Vimshottari Dasha because the model never computed one. The same trick repeats across hundreds of "free AI astrology" apps clogging the App Store charts.

Real astrology needs real astronomy. Real tarot needs a seeded deck. Real Vedic kundli needs sub lord theory. The shortcut you are about to follow gives the model a calculator. The model talks. The calculator computes. You ship a product that does not lie.

This guide is for vibe coders, indie hackers, and absolute beginners. No backend experience required. By the end of the next 30 minutes you will have a live chatbot at your own URL, ready for real users across nine spiritual domains.

What you will actually build

The template is an AI chatbot that handles nine domains from a single subscription, with built in geocoding for any city on top. The user types a question in plain language. The model picks the right calculator behind the scenes, computes the real answer, then explains it in their language. Hindi, Spanish, French, German, Japanese, and more, automatic. No translation switches.

One subscription, one key, one chatbot. No stitching libraries together.

Why the readings stay accurate

The chatbot does not invent positions. Every birth chart, transit, and panchang call goes to a calculation engine verified against NASA JPL Horizons, the same astronomical reference used by space mission planners. Vedic results cross reference DrikPanchang, onlinejyotish.com, and timeanddate.com. Numerology uses the canonical Pythagorean reduction. Tarot draws are seeded so the same user sees the same daily card on the same day, across devices.

That accuracy is what separates this template from the dozens of GPT wrappers selling fake horoscopes on Product Hunt every week. When a real Vedic astrologer tests your bot, the kundli math holds up. The trust layer is built in. Your job is to design a great chat experience on top.

Ready to build it? Try any endpoint live on the API reference to see the data the bot will use, then grab a Starter plan key and the chatbot can be live in the next twenty minutes. The Astrology API product page shows every domain you will cover.

Step 1: Get your RoxyAPI Starter key

This is the engine that powers every reading. One key unlocks all nine domains plus location.

Want to play with the data first? Open the API reference and run any endpoint live in the browser. Daily horoscopes, Vedic kundli, tarot draws, numerology, all of them. No signup, no card. Sanity check the data is real before you commit.

When you are ready to wire the chatbot up, grab the Starter plan on the pricing page. Starter is 39 dollars a month and covers all nine domains plus location under one key. Higher tiers only lift the limits, the domains stay the same.

Drop in your email at checkout. The key arrives in your inbox in seconds. Copy it somewhere safe, you will paste it in Step 5.

Step 2: Get a free Google Gemini key

The chatbot needs a language model to interpret the calculations and chat with users. Google Gemini is the default because it has a generous free tier, fast streaming, and works out of the box.

Visit the Google AI Studio website and sign in with any Google account. Click "Get API Key", then "Create API Key in new project". Copy the key. Save it next to your RoxyAPI key.

The free tier covers fifteen requests per minute and a million tokens per day. Plenty for a personal launch and early users. If you prefer Anthropic Claude or OpenAI GPT, the chatbot supports both with a one line environment variable change.

Step 3: Install Node.js, Git, and VS Code

Three small tools, one time setup. About five minutes total.

Node.js is the runtime that runs the chatbot on your laptop. Visit the Node.js download page and pick the LTS version for your operating system. Click through the installer with default options.

Git is how you download the code. Visit the Git downloads page and grab the installer. On Windows, accept every default. On Mac, Git is usually preinstalled, type git --version in Terminal to check.

VS Code is the editor where you will paste your API keys and tweak the personality later. Download it from the VS Code site. It is optional but makes the rest of the steps easier on the eyes.

Step 4: Clone the chatbot template

Open Terminal on Mac or Linux, or Command Prompt on Windows. Paste these three commands one at a time:

git clone https://github.com/RoxyAPI/astrology-ai-chatbot.git
cd astrology-ai-chatbot
npm install

What just happened: line one downloads the chatbot code into a new folder. Line two moves you into that folder. Line three installs every dependency, about a minute on a normal connection.

The repository is fully MIT licensed. Rename it, rebrand it, charge for it, fork it. No lock in, no hidden fees, no telemetry.

Step 5: Add your API keys

The chatbot reads your keys from a hidden file called .env.local. The repo ships with a template you copy and edit.

Inside the project folder, run:

cp env.example .env.local

Now open .env.local in VS Code. You will see something like this:

ROXYAPI_KEY=your_roxyapi_key_here
LLM_PROVIDER=gemini
GOOGLE_GENERATIVE_AI_API_KEY=your_google_api_key_here

Replace your_roxyapi_key_here with the key you got in Step 1. Replace your_google_api_key_here with the Gemini key from Step 2. Save the file.

That is it. The .env.local file is gitignored by default, so even if you push the project to GitHub later your keys will not leak. Keep this file private and never share screenshots of it on social media or support tickets.

Step 6: Run the chatbot on your laptop

Back in your terminal, type:

npm run dev

After about ten seconds you will see a line like Local: http://localhost:3000. Open that URL in your browser. The chatbot loads with a starry background and an empty chat box, waiting for the first message.

Try a few real prompts:

What does my Saturn placement mean? I was born June 14 1990 in London at 3 am.
Pull a three card tarot spread for my career this week.
What is the meaning of seeing 1111 every day for two weeks?
Calculate my Life Path number. My birthday is July 22 1995.

Each response streams in token by token with real numbers, real card names, and real interpretations. If anything is broken, it is almost always a typo in your .env.local file.

Step 7: Deploy to the public internet for free

You have a working chatbot on your laptop. Now put it on the internet so real users can talk to it.

Click the one click Vercel deploy link. Sign in with GitHub. Vercel forks the repo into your account and asks for two environment variables. Paste the same RoxyAPI key and Gemini key from Step 5. Click Deploy.

Sixty seconds later you have a live URL like your-bot.vercel.app. Free hosting, free SSL, automatic deploys on every push. The Vercel hobby tier handles thousands of users a month at zero cost.

Want a custom domain like astrobot.yourbrand.com? Add it in the Vercel dashboard under Domains, point your DNS, done in ten minutes.

Step 8: Make it yours

The default personality is a warm, knowledgeable spiritual advisor. You can rewrite that in two minutes.

Open src/lib/prompts.ts in VS Code. The system prompt is plain English. Edit it to be mystical, clinical, brand voiced, gen z slangy, Hindi only, whatever fits your audience. Save, push, Vercel redeploys automatically.

To narrow the bot to just tarot, just Vedic, or just numerology, set ROXYAPI_PRODUCTS in your environment to a comma separated list. Fewer tools means faster, more accurate answers. To swap the language model, change LLM_PROVIDER from gemini to anthropic or openai and add the matching key.

Logo, colors, and the star field background live in the src/components and globals.css files. Replace the title, swap the gradient, drop in your brand. The whole UI is Tailwind, no CSS lock in.

How to start charging users

A live AI astrology chatbot is rare, so most founders charge from day one. Three patterns work well.

Free tier with paywall. Give the first three readings free, then gate the chat behind a Stripe link or a Whop product. Users who get one good Vedic reading happily pay 9 to 19 dollars a month for unlimited access.

One off premium readings. Keep daily horoscopes free, charge 5 to 25 dollars for in depth birth chart breakdowns, synastry, or full Celtic Cross spreads.

Embed in an existing community. Drop the chatbot URL into your Discord, Telegram group, or Substack as a member perk. Paid community conversion is much higher when there is a working tool inside, not just content.

At 9 dollars a month per user, you break even on the Starter plan at five paying users. Many founders hit that on launch day.

One last note on AI chatbot pricing

A handful of astrology API providers sell their own bundled "AI chat" endpoint at 0.15 to 0.45 dollars per call, with the model picked for you and responses that can take 8 to 36 seconds. A thousand conversations a month is up to 450 dollars stacked on top of the base subscription, on a model you cannot swap and a price you cannot negotiate.

The template you just deployed flips that. The LLM is your call (Gemini, Claude, OpenAI), the bill is retail (around 0.01 dollars per query at current rates), and the contract is between you and the model vendor directly. When a faster, cheaper model launches, you change one environment variable and ship the upgrade in a minute. No wrapper markup, no model lock in. Owning the model layer is the quiet difference between unit economics that work at one hundred paying users and ones that break.

Things you can ask once it is live

The bot answers in plain language, no special syntax needed:

• "Build a Western birth chart for someone born March 14 1992 at 9 30 am in Cairo."
• "Calculate Gun Milan compatibility between my partner and me. Here are our birth details."
• "Pull a Celtic Cross spread for my career, and explain the King of Cups reversed."
• "Which crystals balance the heart chakra and what does each one do?"
• "Cast an I Ching hexagram for whether I should accept this job offer."
• "What is my Personal Year number for 2026? My birthday is November 5."

Every answer pulls real computed data, then a model wraps it in a personalized response.

Frequently Asked Questions

Q: Do I really need zero coding skills to build an AI astrology chatbot? A: Yes. Every step is copy paste. The only "code" you touch is one environment file with two keys. The repo handles the rest.

Q: How much does it cost to run an AI astrology chatbot in 2026? A: The Gemini free tier plus the 39 dollar RoxyAPI Starter plan plus free Vercel hobby hosting. Total fixed cost: 39 dollars a month, no per call billing surprises.

Q: Is this template free and open source? A: The template is MIT licensed, free forever. You only pay for the RoxyAPI Starter plan and bring your own LLM key. You can fork, rebrand, and resell the chatbot without restriction.

Q: Can the chatbot really do Vedic astrology, not just Western horoscopes? A: Yes. The same chatbot answers questions on Vedic Janam Kundli, Vimshottari Dasha, Gun Milan, KP horary, Panchang, Manglik, and Sadhesati. One bot, both systems.

Q: How do I make sure the calculations are accurate? A: The engine is verified against NASA JPL Horizons and cross referenced against thousands of gold standard tests against authoritative panchang sources. Read the public methodology for exact details.

Ship today, refine tomorrow

Thirty minutes from now, you can have a working AI astrology chatbot live on the internet, covering nine domains, in any language, backed by real astronomy. The path is unglamorous: clone, paste two keys, deploy. The result is unfair: a product most people would charge an agency 10,000 dollars to build.

Grab your Starter plan key, browse the open source starters if you want a different shape (mobile app, Vedic only, tarot only), and try every endpoint live for free on the API reference before you commit. Ship it tonight.