Cloud kitchen India 2026 — the complete cost, licence and setup guide (with real ₹ math)

What a cloud kitchen actually is (30-second version)

A cloud kitchen is a delivery-only food business — a licensed commercial kitchen with no dine-in seating, no walk-in customers and no front-of-house staff. Every order arrives through Swiggy, Zomato, WhatsApp or a direct-order channel. Every order leaves through a delivery rider. The kitchen exists purely to produce food efficiently for a delivery radius of typically 3-5 km.

The model works in India because delivery penetration is deep in every Tier-1 city, commercial rent per square foot is 40-60% cheaper than street-facing dine-in space, and virtual brands let one physical kitchen list on aggregators as multiple restaurants. For a full definition — cloud kitchen vs ghost kitchen vs dark kitchen vs commissary — see the detailed primer at what is a cloud kitchen — 2026 India definition and setup.

The India cloud kitchen market in 2026

India runs roughly 4,000 registered cloud kitchen units nationally in 2026, up from an estimated 1,500 in 2019. Around 65% of the units concentrate in five Tier-1 metros — Mumbai, Bengaluru, Delhi-NCR, Hyderabad and Chennai — because delivery density falls sharply outside these cities and unit economics stop working below roughly 900 orders/month per outlet.

Three-year CAGR sits near 40% by revenue and roughly 25% by unit count. The gap between the two rates matters — new units come in bigger, more multi-brand, and better-capitalised than the 2019-2022 wave. The average new cloud kitchen opening in 2026 runs 4.2 virtual brands from launch and does ₹8-14 lakh monthly revenue by month 6, versus 1.3 brands and ₹3-4 lakh revenue in the 2020 cohort. Directional numbers are drawn from RedSeer, Statista and NRAI industry reports — precise figures vary by methodology.

What has changed most in 2026 is aggregator take-rate. Blended commission across Swiggy and Zomato — including delivery fees, marketing spend, listing fees and payment gateway — now averages 27-32%, up from 22-24% in 2020. Operators who ignored the drift and did not build a direct-order channel are the ones now stuck at 4-6% net margin. Those who added WhatsApp direct ordering early are running 12-16% net margin on the same topline.

The 4 cloud kitchen models — pick before you sign anything

Four operating models dominate India in 2026. Pick the model before you sign a lease — model dictates equipment, staffing and licence tier, and switching later means writing off capex.

Model 1 — Single-brand delivery-only

One kitchen, one brand, one menu. Simplest to operate, cheapest to launch. Best for operators with a strong menu concept and a founder who wants to build a real brand with owned identity, packaging and eventual dine-in expansion. Capex sits at the lower end (₹10-15 lakh). Ceiling is lower too — a single-brand kitchen rarely crosses ₹8 lakh/month in one city.

When to pick: you have a distinctive menu concept and long-term brand ambition. When to avoid: you are testing a market or product-market fit — the ceiling caps your upside.

Model 2 — Multi-brand (virtual brands)

One kitchen, 3-8 virtual brands, one team. Each brand has its own aggregator listing, menu photography and pricing. Prep infrastructure — burners, tandoor, fryer, cold storage — is shared. This is the highest-ROI model in India and is what Rebel Foods, Ghost Kitchens India and every second aggregator-hosted operator run.

When to pick: you want maximum revenue per square foot and are comfortable running the operational complexity. When to avoid: your team is under 4 people or your kitchen equipment is single-cuisine (a pure tandoor kitchen cannot easily add pizza brands).

Model 3 — Aggregator-hosted

You rent a kitchen slot inside a shared facility run by Rebel Foods (Faasos brands), EatClub, Kouzina or Ghost Kitchens India. The host provides fitout, licences, cold storage, delivery integration and often menu-engineering support. You pay 18-28% of revenue as platform fee, on top of Swiggy/Zomato commissions, making blended take-rate 45-55%.

When to pick: you want to test a brand for 6-9 months with sub-₹5 lakh capex risk. When to avoid: you have 3+ year horizon and ₹15+ lakh in capex — your own kitchen becomes cheaper after month 12.

Model 4 — In-restaurant delivery-only expansion

You already run a dine-in restaurant with spare kitchen capacity between meal peaks. You add 2-3 virtual brands operating from the same kitchen — biryani brand in lunch slot, dessert brand at night — using the existing team, licence and gas connection. Marginal capex is ₹50K-2 lakh for photography, packaging and aggregator onboarding.

When to pick: you have a dine-in restaurant already and see capacity idle for 3+ hours daily. When to avoid: your dine-in is already at 85% kitchen utilisation — adding brands will drag ticket time and hurt both channels.

Setup cost — the honest ₹ math (400-600 sq ft, Tier-1 city)

The line items below are the real capex ranges for a single-brand cloud kitchen in a Tier-1 city (Mumbai, Bengaluru, Delhi-NCR, Hyderabad, Chennai) with 400-600 sq ft carpet area. Multi-brand adds ₹5-10 lakh in extra equipment and cold storage.

The two most-underestimated line items are chimney ducting and grease-trap drainage — both are mandated by the Fire NOC and both routinely blow past the fitout estimate. Budget 15-20% contingency on fitout specifically.

City-by-city cost snapshot

Rent, staff cost, licence timeline and break-even months vary meaningfully by city. The table below is a benchmark for a 400-500 sq ft single-brand cloud kitchen doing ₹5-7 lakh/month in revenue by month 6.

Mumbai has the highest revenue ceiling — average ticket size is 15-20% higher and premium brands can charge accordingly — but also the highest cost floor and longest compliance timeline. For a deep dive on Mumbai specifically, including delivery-cluster mapping (Powai, Andheri, Lower Parel, Bandra) and BMC step-by-step, see the Mumbai cloud kitchen cost and setup 2026 playbook.

Licence checklist — India cloud kitchen

Six licences cover a standard India cloud kitchen. Missing any one exposes the operator to aggregator de-listing, ₹1-5 lakh fines under the FSS Act, or municipal sealing.

1. FSSAI State Licence — ₹2,000/year (mandatory). Turnover ₹12 lakh to ₹20 crore, single state. Upgrade to Central Licence at ₹7,500/year if you deliver interstate or list on nationwide e-commerce. Full fee schedule and application walkthrough in the FSSAI license fee 2026 India guide.
2. GST registration. Mandatory above ₹20 lakh turnover; most operators register from day one to reclaim input GST on rent, packaging and utilities. Cloud kitchens fall under 5% GST for restaurants without ITC (composition-like), or 18% with ITC depending on turnover bracket. Details in the GST composition scheme for restaurants guide.
3. Trade Licence. Local municipal registration — BMC (Mumbai), BBMP (Bengaluru), MCD (Delhi), GHMC (Hyderabad), GCC (Chennai). ₹5,000-15,000/year depending on premises size and category.
4. Shops & Establishment. State-level registration under the local Shops & Establishment Act. ₹500-2,000, once. Documents the working hours, weekly off and staff count.
5. Fire NOC. Mandatory because a commercial kitchen uses commercial gas cylinders and high-load electricals. Inspection checks chimney ducting, gas cylinder placement, emergency exit and fire extinguisher count. ₹2,000-5,000 fee plus 2-4 weeks turnaround.
6. Health Trade Licence (state-specific). Applicable in Delhi, West Bengal and a few other states where the local municipal body has notified a separate health licence beyond FSSAI. ₹1,000-5,000, annual.

Unit economics — a Mumbai worked example

A worked P&L clarifies why some cloud kitchens print money and others limp. The example below is a 500 sq ft single-brand cloud kitchen in Powai, Mumbai, doing ₹6 lakh/month gross revenue in month 6. Numbers are directional and match what several operators privately share.

Note the 6% net margin at 28% aggregator commission. The same P&L with 32% commission (no direct-order channel, over-reliance on Swiggy/Zomato ads) drops to break-even. The same P&L with 25% blended commission (25% WhatsApp direct-order share) climbs to 11% net margin — that is the single biggest lever in Indian cloud-kitchen economics right now.

Model the numbers for your own outlet with the restaurant break-even calculator and use the aggregator profit calculator to stress-test different commission blends before you commit.

The tech stack — what a cloud kitchen actually needs

A cloud kitchen tech stack is deliberately thin — three products, roughly ₹7,500/year, running on a single Windows PC and a smartphone.

Ordering Suite (₹199/month)

Consolidates every incoming order — Swiggy, Zomato, WhatsApp direct, QR menu — into one screen. Routes each order to the KDS. Prints KOTs. Manages menu availability, 86-out, price changes across all channels from one place. This is the single most important product in a cloud kitchen — manual Swiggy tablet + Zomato tablet + WhatsApp phone is where 90% of order misses originate. See WhatsApp ordering for how the direct-order channel plugs in.

Desktop POS (₹4,999/year)

Offline-safe billing on any Windows PC. Prints GST-compliant invoices with HSN codes. Manages item-level P&L, sells reports, GST reports for the CA. Works when the internet goes down — critical during a Bengaluru monsoon or a Powai power cut. Product details at products.

Kitchen Display System (KDS)

A tablet or second monitor at the pass showing incoming orders grouped by station (tandoor, chulha, cold-prep, packing). Colour-coded by ticket age — green under 8 minutes, amber 8-15, red 15+. Reduces average ticket time by 20-30% versus paper KOTs and cuts the "where is my order" refund category to near-zero.

Total tech stack cost under ₹7,500/year — recovered in one month of avoided refunds on a ₹6L revenue kitchen. See full pricing at pricing.

Going international — Dubai + GCC pointer

The cloud kitchen model translates cleanly to the GCC — the delivery penetration, the aggregator behaviour, the multi-brand playbook, all identical. The differences are the licence stack (Dubai Municipality + DED trade licence + DCA food code instead of FSSAI + BMC), the currency (AED instead of ₹), and the aggregator mix (Talabat, Noon Food, Deliveroo, Careem in the UAE instead of Swiggy and Zomato).

Unit economics also differ — Dubai average ticket size runs 3-4x Mumbai in AED terms, rent runs 4-6x per square foot, aggregator commission blends around 30-35%. Break-even lands in month 10-14 for a well-run Dubai cloud kitchen. Full founder-level playbook in the how to start a cloud kitchen Dubai 2026 founder guide.

10 common mistakes that kill Indian cloud kitchens

The following 10 mistakes account for roughly 80% of cloud kitchen shutdowns in month 4-9. All are avoidable. All are cheap to fix in month 0 and expensive to fix later.

1. Wrong location — too far from delivery-dense clusters. A cloud kitchen 4 km outside the top 10 delivery pincodes for its city loses roughly 40% of order volume to distance-based aggregator visibility ranking. Signing a lease because rent is 20% cheaper 3 km outside the cluster is a ₹2-3 lakh/month revenue mistake.
2. Under-photographed menu. Aggregator conversion for items with poor photography sits at 2-3% versus 7-9% for professional food photography. The ₹40,000 one-time photography spend is the single highest-ROI capex line item.
3. Manual aggregator-vs-POS reconciliation. Running Swiggy tablet, Zomato tablet and WhatsApp phone separately, and reconciling at end-of-day into a paper KOT book, causes 8-12% order misses. A single Ordering Suite kills this.
4. No WhatsApp direct channel. Kitchens 100% dependent on aggregators run at 28-32% blended commission. Kitchens with 25% WhatsApp direct-order share run at 20-24% blended commission. The 8-point margin swing is the difference between profitable and struggling.
5. Over-broad menu at launch (>25 items). New cloud kitchens launching with 60-80 menu items suffer from prep-time inflation, higher wastage, unclear brand positioning on the aggregator app, and refund spirals. Launch with 15-20 items. Add after month 3 based on order data.
6. Ignoring Wednesday broadcasts (dead daypart). Every cloud kitchen has a dead daypart — usually Tuesday-Wednesday evening or weekday afternoons. A weekly WhatsApp broadcast to opt-in guests fills 8-15% of that slack at zero aggregator commission.
7. No 86-out discipline (refund spiral). Continuing to accept orders for an item you have run out of triggers cancellations, aggregator rating drops (a 4.4 rating falling below 4.2 halves visibility on the app), and refund penalties. Real-time 86-out across all channels is a discipline, not a nice-to-have.
8. Skipping FSSAI State (going Basic when turnover crosses ₹12L). Operators who registered under Basic Registration (₹100/year, up to ₹12L turnover) and then cross the threshold in month 6 face aggregator de-listing when the licence is audited. Register for State Licence from day one if your realistic 12-month projection is above ₹12L.
9. Kitchen designed for dine-in throughput (wrong). Dine-in kitchens optimise for 30-40 minute meal cycles with plating flexibility. Cloud kitchens optimise for 12-18 minute ticket cycles with packaging efficiency. The layout, equipment mix and workflow are different — laying out a cloud kitchen like a dine-in restaurant costs 25-30% throughput.
10. Founder handles ops solo past month 4 (burnout). The founder-cook-manager-marketer model works to month 3. From month 4 the founder becomes the bottleneck — menu additions stall, aggregator issues do not get escalated, staff attrition spikes. Hire a supervisor by month 3 even if margin looks thin — the alternative is closure in month 8.

Frequently Asked Questions

What is the cost of setting up a cloud kitchen in India?

A single-brand cloud kitchen in a Tier-1 Indian city (400-600 sq ft, delivery-dense cluster) sits between ₹10 lakh and ₹27 lakh in one-time capex — kitchen equipment ₹5-12 lakh, fitout ₹3-8 lakh, deposit + first month rent ₹1-3 lakh, initial stock ₹1-3 lakh, POS + tech ₹15-30K, FSSAI + local licences ₹5-20K. Multi-brand kitchens running 3-8 virtual brands from the same square footage need ₹15-35 lakh because equipment stack and cold storage scale up. Monthly recurring costs — rent, staff, gas, aggregator commissions — run ₹2.5-5 lakh depending on city and revenue.

Is a cloud kitchen profitable in India?

Yes, when the unit economics are respected. A Tier-1 cloud kitchen doing ₹6 lakh/month in revenue typically clears an 8-14% net margin — roughly ₹48,000-84,000/month — after food cost of 28-32%, blended aggregator commission of 25-32%, labour 15-18%, rent 8-12% and other opex 5-8%. Kitchens that go below 8% margin are almost always leaking on aggregator commissions (no direct-order channel), over-broad menus (85+ items with high wastage), or under-utilised capacity (single brand instead of 3-4 virtual brands). Break-even for a ₹15 lakh capex outlet is typically 14-20 months at ₹6L monthly revenue.

Which is the best city to start a cloud kitchen in India?

Bengaluru remains the most operator-friendly Tier-1 city — high delivery density in Koramangala, Indiranagar, HSR and Whitefield, ₹35-70/sq ft commercial kitchen rent, aggressive Swiggy/Zomato user base, and 4-6 week licence turnaround. Mumbai has the highest revenue ceiling but rent runs ₹80-180/sq ft in delivery clusters and BMC compliance adds 2-3 weeks. Delhi-NCR (Gurugram, Noida) offers volume plus reasonable rent (₹40-90/sq ft) but staff churn is highest. Hyderabad is the underrated pick — ₹30-55/sq ft rent, growing order density and simpler GHMC compliance. Match city to model: Bengaluru or Hyderabad for a first outlet, Mumbai for premium-pricing brands, Delhi-NCR for scale.

What licenses do I need to start a cloud kitchen in India?

Six licences cover the standard India cloud kitchen. FSSAI State Licence at ₹2,000/year is mandatory (₹7,500/year Central Licence if you deliver interstate or list on nationwide e-commerce). GST registration is required once turnover crosses ₹20 lakh — most operators register from day one to reclaim input GST. Trade Licence from the local municipal corporation (BMC in Mumbai, BBMP in Bengaluru, MCD in Delhi, GHMC in Hyderabad) costs ₹5,000-15,000. Shops & Establishment registration is a state-level requirement, ₹500-2,000. Fire NOC is mandatory for premises using gas cylinders or high-load electricals — timeline 2-4 weeks. A Health Trade Licence (some states, primarily West Bengal and Delhi) applies where the local body has notified it.

How much does a cloud kitchen earn per month?

A single-brand Tier-1 cloud kitchen typically does ₹4-8 lakh in monthly revenue after 3-4 months of operations, once aggregator rating stabilises above 4.2 and repeat-order share crosses 25%. Multi-brand kitchens running 3-4 virtual brands from the same equipment can push monthly revenue to ₹10-18 lakh because they hit multiple cuisine searches on the aggregator app. Revenue per sq ft benchmark: ₹1,200-2,500/sq ft/month for a well-run kitchen. Kitchens stuck at ₹2-3 lakh/month for 6+ months are usually diagnosed as menu-too-broad, photo-quality-poor, or located outside the top 10 delivery pincodes for their city.

Can I run multiple brands from one cloud kitchen?

Yes — the multi-brand model is one of the strongest reasons cloud kitchens work as a business. A 500 sq ft kitchen with 4-5 cooks can operate 3-8 virtual brands simultaneously — biryani brand, pizza brand, healthy bowl brand, dessert brand — each listed separately on Swiggy and Zomato with its own menu, photography and pricing. The economics work because rent, staff and licences are shared across brands, so contribution margin per brand climbs as you add cuisines that share prep infrastructure. Rebel Foods (Faasos, Behrouz, Oven Story) built the largest India business on this model. Guardrails: do not exceed 25 SKUs per brand at launch, and ensure ticket-time stays under 22 minutes across all brands.

What is the difference between a cloud kitchen and a ghost kitchen?

Cloud kitchen and ghost kitchen are used interchangeably in India — both describe a delivery-only kitchen with no dine-in seating and no walk-in customers. The subtle distinction, when operators bother to make one, is that a ghost kitchen usually operates one brand while a cloud kitchen more often runs multiple virtual brands from the same footprint. Internationally the term dark kitchen also appears, and aggregator-hosted kitchens (Rebel Foods, Ghost Kitchens India, Kouzina) are sometimes called host kitchens. From a compliance, tax and unit-economics point of view, all four terms map to the same FSSAI State Licence, the same GST treatment and the same aggregator commission structure.

Do I need a POS system for a cloud kitchen?

Yes — even a delivery-only kitchen needs a POS, primarily to consolidate orders from Swiggy, Zomato and WhatsApp into one screen, print KOTs to the kitchen, and generate GST-compliant invoices. Manual entry from 3-4 aggregator tablets into a paper KOT book is the #1 cause of order misses and refund spirals in cloud kitchens. Online eMenu Ordering Suite at ₹199/month handles aggregator + WhatsApp order aggregation and KDS. Desktop POS at ₹4,999/year handles offline-safe billing and item-level P&L reporting. Stack cost under ₹7,500/year — recovered in one month of avoided refunds.

How long does it take to set up a cloud kitchen?

Realistic end-to-end timeline is 8-12 weeks from lease signing to first Swiggy order. Break-up: premises search and lease negotiation 2-3 weeks, fitout and equipment installation 3-4 weeks, FSSAI State Licence 30-60 working days (run in parallel with fitout), GST and trade licence 1-2 weeks, aggregator onboarding and menu photography 2-3 weeks (also parallel), soft-launch and menu tuning 1-2 weeks. Operators who compress this below 8 weeks usually cut corners on Fire NOC or aggregator photography — both come back to bite in month 2. Operators who stretch past 14 weeks are usually stuck in landlord negotiations or FSSAI query loops.

Should I join an aggregator-hosted cloud kitchen or start my own?

Aggregator-hosted models — Rebel Foods, EatClub, Kouzina, Ghost Kitchens India — take capex to near-zero (₹1-3 lakh security instead of ₹15+ lakh) but charge 18-28% of revenue as platform fee on top of Swiggy/Zomato commissions, which pushes blended take-rate past 45%. Pick aggregator-hosted if you want to test a brand for 6-9 months before committing capex, or if you have menu IP but no operational bandwidth. Pick your own kitchen if you have ₹15-30 lakh capex, a 3-5 year horizon, and the appetite to run multi-brand — the margin math flips decisively in your favour by month 12.