Why is the range so wide?

Because 'custom software' covers everything from a single-screen internal tool to a multi-tenant SaaS with billing, role-based access, integrations, and a mobile app. Three things drive most of the variance: how many distinct user roles the system has, how many external systems it has to talk to (payments, accounting, messaging, government APIs), and how much product judgment the team has to supply versus how much you're handing them. A v1 that you've already specified down to wireframes costs meaningfully less than the same v1 where the team is figuring out the product with you.

What hourly rate should I expect from an Indian software team?

Senior individual freelancers run INR 1,500–4,000 per hour. Mid-tier offshore agencies run INR 2,500–6,000 per hour for blended teams. Boutique studios usually don't quote in hours at all — they quote against a milestone or a launched outcome, with the rate implicit in the price. If a vendor leads with their hourly rate, you are buying time. If they lead with what gets shipped, you are buying a product. Both are valid; just know which one you signed up for.

What's a reasonable budget for an MVP?

For a real MVP — meaning a working product that a real customer could use, not a clickable prototype — budget INR 8L to INR 18L for most categories. You should expect roughly eight to fourteen weeks of build time, one designer's worth of design work, two to four engineers' worth of effort, and a real product owner thinking about scope every week. Anything dramatically cheaper is either a prototype being sold as an MVP, or a team that will run out of budget before the product is usable.

How much should I budget for ongoing maintenance?

Plan for fifteen to twenty-five percent of the initial build cost per year, paid as a monthly retainer. For a INR 20L build, that's INR 25K to INR 40K a month. That covers hosting management, dependency upgrades, small features and copy changes, monitoring and incidents, and the occasional integration that breaks because a third party changed something. Skipping the retainer is the single most common way founders accidentally make their software unusable inside eighteen months.

What hidden costs do most quotes miss?

Five things, in our experience. First, infrastructure and third-party services — hosting, email, SMS, payments, error tracking — usually INR 5K to INR 30K a month from day one. Second, design iteration past the first round, which almost always happens. Third, content and copy work, which is usually scoped to zero and ends up being significant. Fourth, the integration tax: each external system the product talks to adds a meaningful chunk that's easy to under-scope. Fifth, the post-launch tail — the small fixes and decisions in the first eight weeks after going live, which are part of the product, not extra work.

What does custom software actually cost in India in 2026?

The first thing a founder discovers when they ask three Indian software vendors what their idea will cost is that the quotes will not be in the same universe. INR 4L from a freelancer. INR 22L from an offshore agency. INR 35L from a boutique studio. Each comes with a confident PDF and a process diagram. None of them are obviously wrong, and yet they cannot all be right.

The honest answer is that custom software in India in 2026 costs somewhere between INR 6L and INR 40L for a first real build, with most projects landing in a tight band between INR 12L and INR 25L. Above and below that band you are buying genuinely different things. Below it, you are buying time on a single skill. Above it, you are buying multi-quarter capacity on a known plan. The middle is where most "build something real for the first time" projects live, and the middle is what this post is about.

This is the framework we walk a buyer through in a discovery conversation — what drives the number up, what drives it down, what the line items actually are, and what the parts of the budget look like that no one quotes for you.

The four cost archetypes

Before any number, name the shape of the project. Most pricing confusion in this market comes from mixing the four up.

1. Internal tool replacing a spreadsheet

The smallest archetype. One to three user roles, mostly internal, replacing a Google Sheet or an Airtable that a team has outgrown. No payments. Maybe one integration (usually email or WhatsApp). The point of the project is operational leverage, not customer experience.

Realistic range: INR 4L to INR 10L, four to eight weeks of build, one designer-week and one engineer-month equivalent. We write about this archetype in more depth on the internal tools service page; it's the most under-rated category of custom software in India because the ROI is unambiguous and the build is short.

2. Customer-facing v1 of a product

A real product with paying or signing-up customers. Authentication, two to four user roles, a payments integration, an admin surface, and at least one notification channel that has to actually work (transactional email, SMS, WhatsApp). The first launchable version of something the company plans to keep building on.

Realistic range: INR 12L to INR 30L, eight to sixteen weeks, a designer + two to three engineers + a product owner. This is the most common shape we see in the market, and the band where vendor selection matters most. Quotes outside this range for this shape of project are almost always wrong in a way you'll feel later.

3. Multi-tenant SaaS or marketplace

A product with two distinct sides — buyers and sellers, clinics and patients, brokers and developers — or a tenanted product where each customer gets their own isolated workspace. Onboarding flows on both sides. Roles within roles. Reporting. Usually a billing surface with subscriptions or per-transaction fees.

Realistic range: INR 30L to INR 80L, four to eight months, with a real product team. At this size the engagement is no longer a single sprint plan; it's a roadmap with at least two release milestones. Most "we want to build a SaaS" first-time buyers should not start here. They should start with archetype 2 — one side of the market, one customer segment — and grow into archetype 3 after the first version is being used.

4. Internal system at company scale

ERPs, claims systems, multi-branch operations platforms, large back-office automations. Usually a replacement for an aging in-house system or a stitched-together stack of three SaaS products. Long-running, multi-team, high integration count.

Realistic range: INR 50L to a few crore, six to eighteen months, multiple parallel teams. If this is the project you have, the right partner is not a boutique studio and is not a freelancer; it's a senior in-house lead plus an offshore agency or a tier-1 services firm. We write about that decision in the boutique-vs-offshore-vs-freelancer framework.

Most of the rest of this post focuses on the first three archetypes, because that's where the buyer actually has interesting options.

What the line items in a real quote look like

Strip away the cover slides, and a custom software quote in India in 2026 is the sum of roughly seven things. Vendors split them differently, but the underlying math is consistent.

Discovery and product definition. One to three weeks of structured conversation, wireframes, and a written scope. Usually 5–12% of the total. A vendor that skips this and quotes a fixed number off a one-page brief is quoting against their imagination, not yours.

Design. Brand-aware UI design for every screen the team will build, plus a couple of rounds of revisions. For an archetype-2 project, expect 12–20% of the total. Underspending here is a false economy — it's the part of the product the customer actually touches.

Frontend engineering. The visible product. For a modern stack (Next.js / React / a real component system) on a v1, this is typically 25–35% of build cost. More if the product has a heavy mobile surface; we cover that case on the mobile services page.

Backend engineering. APIs, data model, auth, business logic, integrations. Usually 30–40% of the build. The biggest swing factor here is how many external systems are in scope — each integration adds real, non-trivial work that gets under-quoted by inexperienced teams.

Infrastructure and DevOps. Hosting setup, CI/CD, monitoring, environment separation, backups. Smaller than people expect, usually 5–10% of the initial build, but easy to skip and painful to retrofit.

Quality and pre-launch. Manual QA, fixing the issues found, a soft launch with a small group, the bug bash week before going live. Usually 8–12% of the total. Vendors that don't break this out separately are absorbing it into other line items, which usually means it's getting cut when the budget is tight.

Project management and product ownership. The senior person who actually owns the product getting to launch — not just running standups. In a boutique engagement this is usually a partner; in an offshore engagement this is a delivery manager. Either way, it's real time and should be visible in the quote.

For an archetype-2 project at INR 20L, the rough split looks like INR 1.5L discovery, INR 3.5L design, INR 6L frontend, INR 6L backend, INR 1.5L infra, INR 1.5L QA, with PM woven through. Numbers shift, but if a quote is wildly out of this shape, ask why.

Hourly rates are not a useful comparison

The most common mistake first-time buyers make is putting three quotes on a spreadsheet and dividing each by the hours quoted. The freelancer comes out cheapest per hour. The agency comes out in the middle. The boutique studio looks expensive.

The freelancer cannot make product calls, so the missing hours show up later as your hours. The agency can staff capacity but not own the product, so the missing hours show up as a longer timeline and a v1 that needs a v1.5. The studio quotes fewer hours because the team is more senior and writes less throwaway code. The unit being sold is different, and per-hour math punishes the seniors.

The right comparison is what does each vendor commit to delivering, and what do they not. A studio quote that commits to "a launched, customer-usable v1 of the product, with the integrations live, by date X, for INR Y" is not the same product as an agency quote that commits to "a team of three engineers and a designer for sixteen weeks at INR Z." The first is a fixed-outcome bet on the studio. The second is a fixed-capacity bet on you. Both are legitimate; they're just not directly comparable.

The parts of the budget no one quotes

Five categories of cost almost never appear in a first-round quote, and almost always show up by month three. Build them into your number from the start.

Third-party services. Hosting (Vercel, Railway, AWS), transactional email (Postmark, Resend, SendGrid), SMS or WhatsApp (Wati, Twilio, Gupshup), payments (Razorpay, Stripe), error monitoring (Sentry), analytics. For an archetype-2 product, plan on INR 5K to INR 30K per month from day one, scaling with usage. None of this is in the build quote and all of it is mandatory.

Domain, brand, content. A logo, a name, the website copy, the launch email, the in-product copy that the engineer wrote and that you'll want to rewrite. INR 50K to INR 3L is normal, depending on whether you have a brand already.

Iteration past the first design round. Almost every project decides, on seeing the first build, that two or three screens need to be different. This is healthy product development; it is also work that nobody scoped. Reserve 8–12% of the build budget as a design contingency.

Real-world integrations. The integration with Tally that turns out to need a desktop bridge. The GST API that has its own approval timeline. The bank statement parser that has to handle six bank formats. Each of these is a few weeks of work that's easy to under-scope before discovery is done. We've written about the most common of these on the automation service page.

The first eight weeks after launch. Every product needs significant attention in the eight weeks after going live — small fixes, copy edits, a sign-up flow that turns out to confuse users, a report a user actually wanted. This is part of the product, not extra work, and it's where the ongoing retainer earns its keep. Plan for it from day one. Our retainer model is described in more detail here.

A realistic worked example

Take a fairly common shape: a clinic chain with three locations wants a custom system that handles patient intake, appointment booking, doctor scheduling, basic billing, and a partner dashboard for the founders to see usage across locations. No mobile app in v1; a clean responsive web product.

That's an archetype-2 project leaning toward archetype-3. Here's what an honest quote looks like in 2026:

Discovery + product definition: 2 weeks, INR 1.5L
Design (15+ screens, two rounds): INR 3L
Frontend build: INR 6L
Backend build (auth, scheduling, billing, reports, two integrations): INR 7L
Infrastructure + monitoring: INR 1L
QA + soft launch: INR 1.5L
Product ownership through launch: absorbed into above

Total: ~INR 20L, twelve to fourteen weeks.

Then on top of that:

Third-party services: INR 12K/month (hosting, email, WhatsApp, payments, monitoring)
Retainer post-launch: INR 35K/month, scaling up if you start adding the mobile app or a fourth location
Brand and copy work: done already, so INR 0; if not, add INR 1.5L

A vendor offering this same scope for INR 8L is leaving real work out of the quote. A vendor offering it for INR 60L is loading capacity you do not yet need. Either is a yellow flag, not a deal-breaker — but the conversation should be about why their number is so far from the middle.

What drives the number up

Five things, in roughly the order we see them matter:

Number of distinct user roles. Each role is a separate set of screens, permissions, and edge cases. A two-role product is not twice as much work as a one-role product; it's roughly 1.6x. A four-role product is about 3x.
Number of external integrations. Each one is a few weeks. The pain is not the API; it's the testing, the failure modes, the auth tokens that expire, the rate limits. We've written about this in the context of Tally, Zoho, Stripe, and GST integrations (forthcoming).
Mobile native vs. responsive web. Building a real native mobile app roughly doubles the engineering surface area for the same product. A responsive web product that works on a phone is a fraction of the cost.
Compliance and audit requirements. Healthcare, financial services, anything touching regulated data. Adds discovery time, infrastructure cost, and ongoing operational overhead. Worth doing properly, not worth skipping.
Real-time and high-volume features. Live chat, live tracking, real-time inventory across locations, anything that has to update many users at once. Adds infrastructure complexity that small projects don't need.

What drives the number down

Three things, in our experience, matter more than people expect:

A clear, written scope before vendor conversations. A two-page document describing what the product does, who uses it, what the v1 must include, and what is explicitly out of scope is the single highest-leverage thing a buyer can do. It cuts vendor pricing risk and removes the most common cause of overruns.
Saying no to v1 features that can be v2. Most first-time buyers describe a v3 product when asked what they want. The discipline of cutting back to a v1 that ships in twelve weeks instead of twenty-four is worth INR 5L to INR 10L on a typical engagement.
A real product owner on your side. Someone — you, a co-founder, a head of operations — who can make calls within forty-eight hours when the team has questions. Quotes assume this exists. When it doesn't, the project loses one to two weeks per month, and that time gets billed.

For a sense of how this actually plays out across different verticals, our case studies walk through what specific projects ended up costing and why.

How to think about your number

The question is not "how cheap can I get this built." That question almost always produces software that has to be rebuilt within eighteen months, and the second build is more expensive than the first one done right.

The right question is closer to: what is this software worth to my business in the first twelve months after launch? If the honest answer is INR 50L of operational savings or new revenue, an INR 20L build is a clear yes. If the honest answer is INR 5L of saved time, a INR 20L build is a clear no — you should be looking at off-the-shelf SaaS or an internal-tool archetype-1 build for INR 6L.

Most first-time buyers underestimate the value side of this equation and over-optimize the cost side. The vendors who lead with the lowest price know this, and price accordingly. The vendors worth hiring lead with the question of what the software is worth to you, and then back into a number that makes sense for both sides.

If you're trying to set a realistic budget for a custom software build in India in 2026 — or sanity-check a quote you've already received — that's the kind of conversation worth having. Tell us what you're trying to build, and we'll walk through where your project sits in the four archetypes, what we'd quote it at, and where the line items would land.