AI Chatbot SaaS

An AI chatbot for your website, live in one line of code.

Train it on your content, style it to match your brand, and embed it anywhere. Multiple AI models, a real knowledge base, analytics, and honest token pricing — built and hosted by Bitecodes.

Start free Talk to us on WhatsApp

How it works

Step 1

Create a chatbot

Name it, pick a theme, set a welcome message. One account can run many bots.

Step 2

Add your knowledge

Upload PDFs, docs, or a spreadsheet — or paste a URL and we crawl the site. It is chunked, embedded, and indexed automatically.

Step 3

Write the system prompt

Shape its tone and rules with a template and variables like {{company_name}}. Version history lets you restore any earlier prompt.

Step 4

Embed one line of code

Copy the snippet or iframe, paste it into your site, and the assistant is live immediately — answering from your content.

Two ways to embed

A floating widget via a single script tag, or a full-page chat via an iframe. Both respect your domain allowlist and go live the moment you paste them in.

No framework, plugin, or build step
Style-isolated so it never clashes with your site
Desktop, tablet, and mobile — dark and light

Option 1 — JavaScript widget

<script src="https://www.bitecodes.com/widget.js"
  data-chatbot="CHATBOT_ID"
  data-token="PUBLIC_TOKEN">
</script>

Option 2 — iframe

<iframe
  src="https://www.bitecodes.com/chat/CHATBOT_ID"
  width="400" height="600">
</iframe>

Everything a serious chatbot needs

RAG that cites your content

Retrieval-augmented answers grounded in your knowledge base, with optional citations and an honest fallback when it does not know.

Multiple AI models

Choose from the models we enable — GPT, Claude, Gemini, and more — each with its own context window and cost, controlled per plan.

Fully customisable widget

Colours, logo, avatar, position, size, dark and light mode, typing animation, suggested questions — as a floating bubble, popup, or embedded.

Domain protection

Whitelist the domains a bot may run on, wildcards included. Every request is re-checked server-side; unauthorised sites are refused.

Usage & analytics

Chats, token usage, response time, popular questions, geography and devices — with daily, monthly, and yearly charts.

Secure & multi-tenant

Hashed API keys, per-tenant token budgets, rate limiting, audit logs, and a token ledger that caps spend by design.

Simple, token-based pricing

Buy a pack of tokens once and use them whenever you like — there is no subscription and nothing expires monthly. Tokens cover the question and the answer together, so a longer reply costs more than a short one. Answer counts below are approximate for that reason.

Starter

₹499

250,000 tokens · ₹2.00 / 1k tokens

Roughly 500 answers. For one site with steady questions.

Get started

Most chosen

Growth

₹1,499

1,250,000 tokens · ₹1.20 / 1k tokens

Roughly 2,500 answers. For a busy site.

Get started

Scale

₹4,999

6,250,000 tokens · ₹0.80 / 1k tokens

Roughly 12,500 answers. For agencies running several assistants.

Get started

Launch pricing, billed in Indian rupees. Every answer is metered against your balance and recorded in a ledger you can audit, and the assistant stops rather than overspending when a balance runs out.

Frequently asked questions

How do I add the AI chatbot to my website?: Create a chatbot, add your content, then copy the one-line JavaScript snippet or the iframe embed and paste it into your site. The assistant goes live immediately, with no framework or plugin required.
What can I train the chatbot on?: Upload PDF, DOCX, TXT, CSV, Markdown, HTML, or JSON files, crawl your website or import a sitemap, or paste content and build FAQs by hand. Everything is chunked, embedded, and searchable with re-index, update, and delete controls.
Which AI models are supported?: The chatbot is model-agnostic. You choose from the models we enable — such as GPT, Claude, Gemini, DeepSeek, and Qwen — each with its own context window, response limit, and token cost.
How does token pricing work?: Every answer deducts tokens based on the input and output it uses. Buy token packs as you go or take a monthly plan that includes a token allowance. You always see your remaining balance, and the bot pauses rather than overspending.
Is it secure and can I restrict where it runs?: Yes. Each chatbot has a domain allowlist (wildcards supported), API keys are hashed and scoped, requests are rate-limited, and a per-tenant token ledger caps spend. Every admin action is audited.

Let's build something that lasts.

Tell us what you're working on. We'll reply within one business day with thoughts on how we can help.

Start a project View our work

How it works

Step 1

Create a chatbot

Name it, pick a theme, set a welcome message. One account can run many bots.

Step 2

Add your knowledge

Upload PDFs, docs, or a spreadsheet — or paste a URL and we crawl the site. It is chunked, embedded, and indexed automatically.

Step 3

Write the system prompt

Shape its tone and rules with a template and variables like {{company_name}}. Version history lets you restore any earlier prompt.

Step 4

Embed one line of code

Copy the snippet or iframe, paste it into your site, and the assistant is live immediately — answering from your content.

Everything a serious chatbot needs

RAG that cites your content

Retrieval-augmented answers grounded in your knowledge base, with optional citations and an honest fallback when it does not know.

Multiple AI models

Choose from the models we enable — GPT, Claude, Gemini, and more — each with its own context window and cost, controlled per plan.

Fully customisable widget

Colours, logo, avatar, position, size, dark and light mode, typing animation, suggested questions — as a floating bubble, popup, or embedded.

Domain protection

Whitelist the domains a bot may run on, wildcards included. Every request is re-checked server-side; unauthorised sites are refused.

Usage & analytics

Chats, token usage, response time, popular questions, geography and devices — with daily, monthly, and yearly charts.

Secure & multi-tenant

Hashed API keys, per-tenant token budgets, rate limiting, audit logs, and a token ledger that caps spend by design.

Simple, token-based pricing

Starter

₹499

250,000 tokens · ₹2.00 / 1k tokens

Roughly 500 answers. For one site with steady questions.

Get started

Most chosen

Growth

₹1,499

1,250,000 tokens · ₹1.20 / 1k tokens

Roughly 2,500 answers. For a busy site.

Get started

Scale

₹4,999

6,250,000 tokens · ₹0.80 / 1k tokens

Roughly 12,500 answers. For agencies running several assistants.

Get started

Launch pricing, billed in Indian rupees. Every answer is metered against your balance and recorded in a ledger you can audit, and the assistant stops rather than overspending when a balance runs out.

Frequently asked questions

How do I add the AI chatbot to my website?

Create a chatbot, add your content, then copy the one-line JavaScript snippet or the iframe embed and paste it into your site. The assistant goes live immediately, with no framework or plugin required.

What can I train the chatbot on?

Upload PDF, DOCX, TXT, CSV, Markdown, HTML, or JSON files, crawl your website or import a sitemap, or paste content and build FAQs by hand. Everything is chunked, embedded, and searchable with re-index, update, and delete controls.

Which AI models are supported?

The chatbot is model-agnostic. You choose from the models we enable — such as GPT, Claude, Gemini, DeepSeek, and Qwen — each with its own context window, response limit, and token cost.

How does token pricing work?

Every answer deducts tokens based on the input and output it uses. Buy token packs as you go or take a monthly plan that includes a token allowance. You always see your remaining balance, and the bot pauses rather than overspending.

Is it secure and can I restrict where it runs?

Yes. Each chatbot has a domain allowlist (wildcards supported), API keys are hashed and scoped, requests are rate-limited, and a per-tenant token ledger caps spend. Every admin action is audited.