Google.org and The/Nudge leaders at the New Delhi launch of the $15 Mn MSME AI programme
NewsAI & DeepTechFintech

Google.org Commits $15 Mn to The/Nudge to Expand Credit Access for Indian MSMEs

The money funds AI tools for credit, compliance and market access, with an open call for proposals built into the programme.
by Aalam Rohile
3 min read

SUMMARY

  • Google.org has committed $15 Mn to The/Nudge to build AI tools that help Indian MSMEs access formal credit, find markets and handle compliance.
  • It’s a grant to a nonprofit, not a loan programme, and an open call for proposals will source solutions.
  • Founders building lending, compliance or MSME-tech products should watch the open call at Samriddhi by The/Nudge for openings.

Google.org has committed $15 Mn to The/Nudge, the Bengaluru nonprofit formerly known as The/Nudge Institute, to build AI tools for India’s micro, small and medium enterprises. The announcement came in New Delhi on 14 October 2026.

The headline sounds like a lending programme. It isn’t. This is a grant to fund technology that could make formal credit easier for small businesses to reach, and that distinction matters if you’re building anything in this space.

What the $15 Mn is meant to build

According to Inc42, the funding will back scalable AI tools that help small businesses improve access to formal credit, discover new markets and meet compliance requirements. An open call for proposals is part of the plan, and MediaNews4U reports it’s listed on Samriddhi by The/Nudge.

There’s a second piece. A multilingual AI platform will train people in Tier III and Tier IV towns in job-ready skills and connect them with small businesses that need those skills.

Preeti Lobana, VP and country manager at Google India, said Google is “putting open digital tools directly into the hands of small business owners.” She also said the programme will back stage-gated grand challenges, because compliance, market access, quality and credit problems reinforce each other.

The initiative also aims to set open standards that help Indian MSMEs become more competitive, creditworthy and visible in global markets.

Why credit sits first on the list

Google’s own estimate is that more than 90 million small businesses in India employ over 345 million people. Yet MSME credit penetration is only 14%, against 37% in China and 50% in the US, according to an EY report cited by Policy Circle.

Estimates of the funding gap vary. A parliamentary standing committee put it at ₹20-25 lakh crore, while a Mavenark report, as Outlook Business reported it, estimated about ₹28 lakh crore. Different methods, so read it as a range rather than a single number.

The reasons are familiar. Many small firms have no credit history or collateral, and many operate informally without clean repayment records. That’s why data-based lending is catching on. Public sector banks sanctioned 98,995 MSME loan applications under digital credit assessment between 1 April and 15 July 2025, YourStory reports.

ShuruKar launched in Patna: What It Is, Who Can Join and How It Works
ShuruKar launched in Patna: What It Is, Who Can Join and How It Works
YourStory's The Bharat Project, working with DPIIT and Startup India, wants young Indians to turn a problem in their own district into a business. Here is what …

It isn’t Google’s first $15 Mn MSME programme

In 2021, Google announced $15 million in debt capital with SIDBI for micro enterprises with turnover up to ₹5 crore, with loans of ₹25 lakh to ₹1 crore, as Entrackr covered at the time.

That was lending capital. This one funds tools and a talent platform. Same number, different mechanism, and easy to mix up in headlines.

Who builds it, and what we don’t know yet

This is still a developing story, so here’s what’s confirmed and what isn’t.

StartupTalky reports that although the grant names The/Nudge, the programme will be built by the^delta, which it describes as The/Nudge’s social entrepreneurship arm under a new name. It also says the^delta already runs an AI-for-impact programme with Meta. Other outlets name only The/Nudge, so treat the^delta detail as one outlet’s reporting for now.

StartupTalky puts the grant at about ₹143 crore and cites Forbes India for The/Nudge’s annual budget at around ₹100 crore. The announcement doesn’t say how many years the money will be deployed, so $15 Mn shouldn’t be read as one year’s spend.

Eligibility, deadlines and grant sizes for the open call also weren’t detailed in the coverage we reviewed.

What this means for founders

India already has digital lenders built around MSMEs. Inc42 names Indifi, Techfino, Lendingkart, FlexiLoans and Aye Finance among those using technology to lend to this segment. Their edge has mostly been data and underwriting speed.

Here’s our read: open standards and shared tools could lower the cost of the data work every lender repeats. That’s StartupIndiaX’s view, not a claim from Google or The/Nudge.

The upside Google is pointing at is large. A World Economic Forum report from October 2025 estimated AI adoption could unlock $490-685 Bn in economic value for India’s MSME sector, Inc42 notes.

Google also has an equity route for startups. Its AI Futures Fund and Accel’s Atoms programme announced co-investments of up to $2 million per pre-seed Indian AI startup in November 2025, per Storyboard18. This grant is a different channel, aimed at the MSME problem itself rather than startup equity.

Read More: TAM vs SAM vs SOM: The Formula Every Indian Founder Gets Wrong, Explained on One ₹100 Crore Market

What to watch next

  • Track the open call on Samriddhi by The/Nudge, since eligibility, deadlines and grant sizes haven’t been published in the coverage so far.
  • Watch for the deployment timeline, because the $15 Mn has no stated duration and could be spread over several years.
  • See whether the open standards turn into credit and data tools that lenders and MSMEs actually adopt.

Building for MSME credit or compliance? Tell us in the comments what you’d want from an open call, and explore more funding coverage on StartupIndiaX.

FAQs

What did Google.org announce for Indian MSMEs?

Google.org committed $15 Mn to The/Nudge to fund AI tools for MSME credit, market discovery and compliance. The programme also includes an open call for proposals and a multilingual platform for skills training in Tier III and Tier IV towns.

Is this a loan programme for small businesses?

No. It’s a grant to a nonprofit to build technology and open standards. Google’s separate 2021 programme with SIDBI was lending capital, with loans between ₹25 lakh and ₹1 crore for micro enterprises.

Who is The/Nudge?

The/Nudge is a Bengaluru-based nonprofit, formerly The/Nudge Institute, founded in 2015 by Atul Satija. It works on livelihoods and economic inclusion in India and runs incubator, accelerator and prize programmes.

Can startups apply for the funding?

An open call for proposals is part of the programme, listed on Samriddhi by The/Nudge. Eligibility, deadlines and grant sizes weren’t detailed in the coverage reviewed, so check the portal directly before preparing an application.

How big is India’s MSME credit gap?

Estimates differ. A parliamentary standing committee put the gap at ₹20-25 lakh crore, while Mavenark estimated about ₹28 lakh crore. Credit penetration in the sector is just 14%, according to an EY report.

0 comments 17 views
FacebookTwitterLinkedinWhatsapp
TAM vs SAM vs SOM The Formula Every Indian Founder Gets Wrong, Explained on One ₹100 Crore Market
Startup LearningFounder Toolkit

TAM vs SAM vs SOM: The Formula Every Indian Founder Gets Wrong, Explained on One ₹100 Crore Market

One ₹100 crore market, taken from TAM to SAM to SOM in three steps, so you can see exactly where the number shrinks and why the last line can't be a guess.
by Aalam Rohile
3 min read

SUMMARY

  • TAM is customers times price, SAM filters it for fit, and SOM applies a win rate built from your actual funnel.
  • In the worked example, a ₹100 crore TAM becomes a ₹40 crore SAM and a ₹2 crore SOM, or 5% of SAM.
  • Treat the SOM percentage as a result, not an input. Work backwards from paying customers, trials and reach.

The first line of the TAM SAM SOM formula is easy. TAM is customers times price. The chain usually breaks after that.

SAM turns into a vague “we’ll start with the metros”, and SOM becomes “1% of the market”, a number with no plan behind it. Investors notice, because that’s the point where the maths stops being maths.

Here’s the TAM SAM SOM formula in three lines, followed by one ₹100 crore market worked from top to bottom. The startup is an illustration, not a real company, and every input is an assumption you should replace with sourced data for your own business.

The three lines to remember

Every guide words it slightly differently, but this is the version most investor guides use:

  • TAM = potential customers × annual revenue per customer. Everyone who could ever buy, if you won all of them. Wise’s guide lays out the same basic formula.
  • SAM = TAM × the share that fits your product, price and geography. The slice you can actually serve today.
  • SOM = SAM × a realistic win rate. What you can capture in roughly three to five years, per the same guide.

That third line is where TAM vs SAM vs SOM stops being a vocabulary lesson. A win rate isn’t a mood. It has to come from your team size, your budget and your funnel. For plain-English definitions of each term, TechTarget’s explainer is a good reference.

Read More: TAM SAM SOM Examples for Indian Startups: 7 Realistic Scenarios

Taking one ₹100 crore market all the way down

Imagine an invoicing app for freelancers and solo service providers in India, priced at ₹1,000 a year (assumption).

Step 1: TAM. Assume 10 lakh freelancers could use it. 10 lakh × ₹1,000 = ₹100 crore.

Step 2: SAM. Assume 40% of them invoice enough Indian clients, and are GST-registered or about to be, to need the app. You also only support Android, English and Hindi for now. TAM × 40% = 4 lakh customers, or ₹40 crore.

Step 3: SOM. Don’t write “5% of SAM” yet. Work backwards from what you can actually do.

Say you want 20,000 paying customers in three years. If one in three trials converts (assumption), you need 60,000 trials. If one in five visitors starts a trial (assumption), you need 3 lakh visitors, about 1 lakh a year. Then 20,000 × ₹1,000 = ₹2 crore.

Only now do you divide. ₹2 crore ÷ ₹40 crore = 5% of SAM. The percentage is the result, not the input.

The full chain reads ₹100 crore, then ₹40 crore, then ₹2 crore. Every step has a reason you can say out loud.

Read More: StartupIndiaX free Founder Toolkit for Indian startups

Three places the chain breaks

TAM equals SAM. If both numbers match, you haven’t done the filtering. It’s also tempting to grab a big official count and call it your market. IBEF reported over 7.83 crore enterprises on the Udyam platforms till February 2026, per the Ministry of MSME data. Multiply that by ₹1,000 and you get a ₹7,830 crore TAM. But a registry is a headcount, not a buyer list.

SOM as a fixed percentage. “We only need 1%” tells an investor nothing about how you’ll win that 1%. A bottom-up SOM, built from sales capacity, win rate and deal size, connects the slide to your operating plan.

Mixed units. Customers, revenue and GMV get swapped halfway through the slide. Pick one unit, revenue per year, and keep it across all three lines.

Read More: How to raise startup funding in India

Is a ₹100 crore market big enough to raise on?

On its own, probably not for venture funding. Vendor guides such as Prospeo’s cite rough venture-scale thresholds: a TAM above $1 billion, a SAM above $100 million and a SOM above $10 million. Treat those as folk benchmarks, not rules. Against them, ₹100 crore looks small.

That doesn’t make it a bad business. A founder with this market has two honest paths. Build a profitable niche company, or widen the product so the TAM grows. Either way, say which. eChai’s advice for Indian founders is to state your assumptions plainly rather than hide behind the acronyms.

Read More: TAM SAM SOM Calculator

What to carry into your deck

  • Write the chain as three lines: customers × price, a fit filter, then a funnel-based win rate.
  • Back-solve SOM from reach, trials and conversion. If you can’t, the percentage is a guess.
  • Label every input as an assumption or a sourced figure before the deck leaves your laptop.

Ready to try it on your own market? Run your numbers in the free TAM SAM SOM Calculator, then tell us in the comments which line of the formula trips you up most.

FAQs

What is the TAM SAM SOM formula?

TAM is potential customers times annual revenue per customer. SAM is TAM multiplied by the share that fits your product, price and geography. SOM is SAM multiplied by a realistic win rate, ideally built from your funnel and sales capacity.

What is the difference between TAM, SAM and SOM?

TAM is the total demand if you won every customer. SAM is the part you can serve with your current product, price and geography. SOM is the part you can realistically win in the next few years.

How do you calculate SOM from SAM?

Multiply SAM by your realistic win rate, then check it bottom-up. Work backwards from paying customers to trials to reach. If your funnel can’t deliver that many customers, the percentage is too high.

Is SOM always a percentage of SAM?

It’s expressed as one, but it shouldn’t start as one. Build the customer count from your funnel and budget first, then divide by SAM. In the ₹100 crore example, that came to 5%.

Can TAM and SAM be the same number?

Rarely. If they match, you probably haven’t filtered for product fit, price, language or geography. It can happen when a product truly serves the whole market, but investors will expect you to prove it.

What is a good TAM SAM SOM ratio?

There’s no universal ratio. Some vendor guides cite a TAM above $1 billion for venture-scale businesses, but those are folk benchmarks. A SOM you can trace to a funnel matters more than any ratio.

0 comments 19 views
FacebookTwitterLinkedinWhatsapp
Founders sizing their market on a whiteboard with three concentric circles showing TAM SAM SOM in rupees
Founder ToolkitStartup Learning

TAM SAM SOM Examples for Indian Startups: 7 Realistic Scenarios

Seven worked examples in rupees, from small-shop software to gig-worker fintech, showing how to turn a giant market number into one you can defend in front of an investor.
by Aalam Rohile
3 min read

SUMMARY

  • Build SOM bottom-up from customers, price and the cash you can spend. Treat TAM as context, never as the headline number.
  • Across these seven scenarios, SOM lands between 0.5% and 3% of SAM, and every one depends on distribution, not demand.
  • Use Indian data (Udyam, Vahan, kirana and gig-worker counts) and label every unsourced input as an assumption before an investor does.

Open enough pitch decks and the market slide starts to look identical. One big number, one tiny percentage, and a promise that “even 1%” makes the company worth funding.

Investors have seen that slide too many times. In India it’s worse, because “1.4 billion people” isn’t a customer base. One framing from the Indus Valley Report, summarised by eChai, puts India’s real spending market (India1) at roughly 100 million people. Size your SAM from the wrong end and the whole slide is fiction.

So here are seven TAM SAM SOM examples for Indian startups, each worked out in rupees. A few ground rules first. Every startup below is an illustrative scenario, not a real company. Numbers with a source link are published figures. Anything marked assumption is a placeholder you should replace with sourced data for your own business.

The method stays the same each time. TAM is everyone who could ever buy. SAM is the slice your product, price, language and geography can serve. SOM is what you can win in about three years, built from customers and cash, not from a percentage.

1. A billing app for small shops: 7.83 crore registrations aren’t 7.83 crore customers

Say a founder in Pune is building billing and inventory software for small retailers. The Ministry of MSME says over 7.83 crore enterprises were registered on the Udyam platforms till 28 February 2026, as IBEF reported. That count includes the Udyam Assist platform, which covers informal micro units, so it’s a headcount, not a buyer list.

Multiply 7.83 crore by ₹6,000 a year (assumption) and the TAM is ₹46,980 crore. Now filter.

  • SAM: Assume 15% are retailers or traders who bill customers daily, about 1.17 crore shops. Assume you can sell and support in only eight states, covering 25% of them. That’s 29.4 lakh shops, or ₹1,762 crore.
  • SOM: Assume 10 field reps each close 70 shops a month. That’s roughly 8,400 shops a year, or about 25,000 in three years before churn. 25,000 × ₹6,000 = ₹15 crore, or 0.85% of SAM.

The SOM comes from a hiring plan, not from a percentage. That’s what makes it believable.

2. The kirana trap: GMV isn’t TAM

Now picture a restocking app for neighbourhood stores. Industry estimates put kirana stores at more than 12 million, accounting for about 90% of grocery sales, according to Flourish Ventures. Companies like 1K Kirana, which pools store demand and negotiates with manufacturers, have built on exactly this thesis.

Here’s where founders slip. Assume each kirana buys ₹1 lakh a month wholesale (assumption), or ₹12 lakh a year. Across 1.2 crore stores that’s ₹14.4 lakh crore of goods. But that’s GMV. If you earn a 1.5% take rate (assumption), your TAM in revenue is ₹21,600 crore.

  • SAM: Assume 8% of stores sit in the ten cities where you can deliver next day. That’s 9.6 lakh stores, ₹1,728 crore in revenue.
  • SOM: 6,000 active stores buying ₹12 lakh a year is ₹720 crore of GMV, and ₹10.8 crore of revenue.

Show both numbers on the slide, and label which is which. Investors will ask anyway.

3. A D2C skincare brand, where cash caps the SOM

Take a skincare brand for oily skin, sold online to Tier 1 shoppers. No single published number fits this cleanly, so every input below is an assumption. Pull yours from a recognised industry report and cite the year.

Assume 2 crore online beauty buyers, each spending ₹1,500 a year (₹600 an order, 2.5 orders). TAM: ₹3,000 crore.

  • SAM: Assume 30% have the oily-skin concern and 40% live in metros and Tier 1 cities. That’s 24 lakh buyers, or ₹360 crore.
  • SOM: Assume a ₹1.5 crore marketing budget and a ₹500 customer acquisition cost. That buys 30,000 customers, and 30,000 × ₹1,500 = ₹4.5 crore, or 1.25% of SAM.
Concentric circle diagram showing a D2C skincare brand's TAM of 3,000 crore rupees, SAM of 360 crore and SOM of 4.5 crore TAM SAM SOM Examples

This is a serviceable obtainable market example where money, not demand, sets the ceiling. It also assumes the full year’s spend lands in the year you win the customer, so trim it if buyers join late.

4. A battery-care plan for e-scooters: a market that moves every month

Electric two-wheelers are the cleanest data story in this list. Vahan figures show 14,01,663 e-2Ws were retailed in FY2026, up 22% year on year. EV penetration in two-wheelers reached 11.2% in July 2026.

Imagine a battery-care plan priced at ₹5,000 (assumption), sold to each new buyer. About 14.0 lakh buyers a year makes the TAM ₹700 crore. Notice it counts a yearly flow of new buyers, not the whole fleet on the road. Pick one and say which.

  • SAM: Assume 40% of buyers are in your three states and 35% of those aren’t tied to an OEM service package. That’s 1.96 lakh buyers, ₹98 crore.
  • SOM: Assume dealer tie-ups convert 3%. That’s 5,880 plans and ₹2.94 crore.

Tie your SOM to your own dealers, not to a brand’s market share. Shares swing hard: Ola Electric’s e-2W sales fell from 3,44,009 in FY25 to 1,64,295 in FY26, according to Vahan-based tallies. Re-run this model every quarter.

Read More: Your TAM Slide Is Probably Wrong: How to Calculate TAM, SAM and SOM the Way Indian VCs Expect

5. Gig-worker fintech: size it through the platform that owns the user

Consider accident cover plus a small instant advance for delivery and ride partners. The Economic Survey 2025-26 puts gig and platform workers at 1.2 crore in FY25. NITI Aayog projected 2.35 crore by 2029-30, up from 77 lakh in 2020-21.

Size on today’s number, not the projection. Assume ₹1,200 of revenue per worker a year (assumption). TAM: 1.2 crore × ₹1,200 = ₹1,440 crore.

  • SAM: Assume 30% work on platforms you can partner with. That’s 36 lakh workers, ₹432 crore.
  • SOM: Assume the partner funnel converts 2%. That’s 72,000 users and ₹8.64 crore.

The catch is that one partnership decides everything. If the platform says no, the SOM is zero. Say that out loud in the deck.

Read More: StartupIndiaX free Founder Toolkit for Indian startups

6. The paid learning app and the 1.4 billion mistake

A founder building a skills app at ₹2,400 a year (assumption) is tempted to start at 140 crore people. That gives a TAM of ₹3.36 lakh crore, and it’s nonsense.

Start from India1 instead. Use roughly 10 crore people as the honest TAM: ₹24,000 crore. Then filter.

  • SAM: Assume 25% are English-speaking working professionals, and 10% of those will pay for skill apps. That’s 25 lakh people, or ₹600 crore.
  • SOM: Assume 40,000 subscribers by year three. That’s ₹9.6 crore, or 1.6% of SAM.

The ₹3.36 lakh crore version isn’t just bigger. It tells an investor you haven’t asked who can pay.

Read More: How to raise startup funding in India

7. Built in Bengaluru, sold in dollars: the global-from-India split

Last, a support-ticket tool for online store owners abroad. eChai’s advice for founders is to split India-domestic from global-from-India, because they’re two different businesses with different pricing and different investors.

Assume 20 lakh target stores in English-speaking markets, a price of $300 a year, and ₹90 to the dollar (all assumptions). That’s ₹27,000 per customer and a TAM of ₹5,400 crore.

  • SAM: Assume 15% sell enough to need a helpdesk. That’s 3 lakh stores, ₹810 crore.
  • SOM: Assume 1,500 paying merchants in three years. That’s ₹4.05 crore, or 0.5% of SAM.

Add exchange-rate movement to your assumptions list. It can change this SOM without a single new customer.

Read More: Startup Runway Calculator

What the seven have in common

Look back at the numbers. SOM never passes 3% of SAM, and none of them were chosen first. Each came from a sales team, a budget, a dealer network or a partner. That’s the difference between top-down vs bottom-up market sizing: top-down gives you context from industry reports, while bottom-up gives you a number you can defend line by line.

Bar chart of SOM as a share of SAM across seven Indian startup scenarios, ranging from 0.5 percent to 3 percent

Build your market size slide for a pitch deck with SOM as the hero number and TAM above it as context. Then check that your SOM can be funded. The Startup Runway Calculator shows how many months of cash you have to reach it, and the Funding Round Dilution Calculator shows what raising for it costs you.

Read More: Funding Round Dilution Calculator

What to take into your next pitch

  • Start with SOM and build it from customers, price and budget. Use TAM as context, not as the headline number.
  • Mark every unsourced input as an assumption, and attach a year and source to every published figure.
  • Re-run your sizing each quarter. Policy, pricing and market share can move your SAM faster than your sales do.

Ready to size your own market? Run your numbers in the free TAM SAM SOM Calculator, then tell us in the comments which of these seven looks closest to your business.

FAQs

What is a TAM SAM SOM example?

Take a billing app for small shops. TAM is every registered Indian business that could pay for it. SAM is the retail and trading units you can actually sell to. SOM is the number of shops you can sign in three years.

What is the difference between TAM and SAM?

TAM is the whole demand if you won every customer. SAM is the part your product, price, language and geography can serve today. A huge TAM with a tiny SAM usually points to a go-to-market problem.

How do you calculate TAM, SAM and SOM?

Multiply customers by annual revenue per customer for TAM. Apply filters for fit and reach to get SAM. Then build SOM from the customers you can realistically sign with your team, channels and budget.

What SOM percentage is realistic for a startup?

There’s no fixed rule. In the seven scenarios above, SOM lands between 0.5% and 3% of SAM, but each figure comes from a sales or budget plan. Treat any percentage as a sanity check, not a method.

How do I find TAM data for Indian markets?

Start with official sources: the Udyam portal and IBEF for MSMEs, the Vahan dashboard for vehicles, NITI Aayog and the Economic Survey for labour data, and listed-company filings. Note the year of every number you use.

How do I calculate market size without expensive reports?

Count customers yourself. Use government registries, trade bodies, marketplace listings and 20 to 30 customer calls to test price. A bottom-up model built this way is usually easier to defend than a paid report.

Why do investors ask for TAM, SAM and SOM?

They want the ceiling, the slice you can serve and the near-term path in one view. The logic connecting the three layers tells them more about the founder than the headline number does.

Is top-down or bottom-up market sizing better?

Use both. Top-down gives context from industry reports, while bottom-up builds from customers, price and capacity. Investors tend to trust the bottom-up number and use the top-down one as a cross-check.

How big should a TAM be to raise venture funding?

Rules of thumb circulate, such as a TAM above ₹10,000 crore for venture-scale returns. They’re folk benchmarks, not rules. The ratio between TAM and a defensible SOM matters more.

How often should I update my TAM, SAM and SOM?

Update whenever a driver moves: a price change, a new city, a policy shift or a quarter of fresh sales data. In fast markets like electric two-wheelers, re-run it every quarter.

What if my startup has no sales history?

Use comparable startups, pilot results and customer interviews to set price and conversion assumptions. Label them as assumptions, and update them after your first 90 days of selling.

0 comments 20 views
FacebookTwitterLinkedinWhatsapp
Alexandr Wang and Meta's Muse Wonder launch graphic announcing the Muse Wonder and Muse Wonder Full AI models
AI & DeepTechNewsTechnology

Muse Wonder Explained: Meta’s First Frontier AI Model Costs $4 Per Million Tokens

Meta's Muse Wonder and Muse Wonder Full landed on October 12 with frontier claims, a paywall around the stronger model, and a price tag founders building AI agents can't ignore.
by Aalam Rohile
3 min read

SUMMARY

  • Meta launched Muse Wonder and Wonder Full on October 12, with API access at $4 per million input tokens and $20 per million output tokens.
  • Wonder Full, the stronger model, is limited to Muse Code subscribers on $50 or $200 plans and isn’t offered through the API.
  • Benchmark claims come from Meta’s own testing, so founders should run their own evals before moving agent workloads over.

Meta spent 14 months rebuilding its AI operation, and on October 12 it showed the first big result. Muse Wonder and Muse Wonder Full are the company’s first attempt at a frontier model, released by Meta Superintelligence Labs, TechCrunch reported.

Alexandr Wang, who leads Meta’s AI efforts, announced the launch on X. Here’s his post:

What Muse Wonder actually is

Muse Wonder is the bigger sibling of Muse Spark. Spark was built to be lightweight and quick. Wonder is larger, and Meta says it’s built for personal and professional agentic tasks.

Agentic just means the AI does the work instead of only answering questions. Meta’s examples include working through an insurance claim, making sense of a lease, and building a working app.

Two features carry most of the pitch. Active Memory lets the model carry details across conversations. Active Focus keeps it going when a task gets hard, so it tries other approaches, recovers from errors, and checks in before anything important.

On the technical side, OpenRouter’s listing shows a 1 million token context window and support for text, images, video, audio and PDFs. Meta trained the models at its Prometheus data center in Ohio.

Read More: Meta Muse Charm: The Keychain AI Gadget That Could Change How India Uses Apps

Two models, two doors

Here’s where it gets awkward for anyone hoping to try the best version. Muse Wonder is available through the Meta Model API and Muse Code, and it’s headed to the Muse app soon, starting with subscribers.

Muse Wonder Full, the stronger one, is only for Muse Code subscribers. It isn’t served through the API at all.

CNBC reported what entry costs. Muse app users need a $20 or $100 monthly plan to get Wonder, while Wonder Full sits on Muse Code’s $50 or $200 plans. Inside Meta the model went by the codename Watermelon, a joke Wang kept alive with a fruit emoji in his launch post.

Comparison of Muse Wonder and Muse Wonder Full showing availability through the API, Muse Code and Muse app, with monthly plan prices

The $4 question

On the API, Wonder costs $4 per million input tokens and $20 per million output tokens. TechCrunch puts that roughly level with GPT-6 Sol or Opus 5.5.

It’s also a big jump from Meta’s own earlier model. OpenRouter lists Muse Spark 1.3 at $1.25 and $4.25 for the same units, so Wonder costs about three times as much on input and nearly five times as much on output.

Wang told CNBC the pricing would stay “meaningfully below” the leading models from OpenAI and Anthropic. That’s hard to square with a price sheet that matches Opus 5.5, and nothing we found from Meta explains the gap.

Bar chart comparing Muse Spark 1.3 and Muse Wonder API prices per million input and output tokens

Meta’s scorecard, and why to hold the applause

In Meta’s own testing, Wonder and Wonder Full score comparably to Fable 5.1 and GPT-6 Astra. Meta also says Wonder beats Fable, Opus and Astra on MemoryBench, a test of how well models remember. Wang went bigger on X, calling it the best model in the world for personal agents and professional agents.

Those are Meta’s numbers. Interesting Engineering noted that the announcement gave no numerical scores in the material it reviewed, which makes any gain hard to size. Meta says a full safety report is coming soon.

What this means for Indian founders

Meta’s announcement doesn’t spell out regional availability, and the Muse app launched in the US first, on September 8. Indian users shouldn’t count on the consumer app showing up soon.

The API is a different story. It’s a developer product priced in dollars, and Indian startups building agents for support, sales or back-office work already weigh model bills against thin margins. Wonder gives them one more option to test, not a reason to switch.

Treat it like a D2C brand treats a new ad channel. Test small, measure cost per outcome, then decide whether to scale. It also keeps the pressure on homegrown model builders such as Sarvam AI to show where they win, whether that’s price, language or fit for Indian use cases.

Read More: Did Alexandr Wang’s Meta Muse Justify a $14.3 Billion Bet? It Beat ChatGPT, Then the Stock Slipped

The real test is who pays

CNBC framed the launch as a test of whether people will pay for AI, and that’s the sharper reading. The Muse app is free, with subscriptions for heavier use, and Meta wants AI revenue beyond its ad business. Wonder puts its strongest models behind paid plans.

It’s also the first frontier-class release since Wang restructured Meta’s AI effort in August 2025, per TechCrunch. Free got Meta users in the door. A $20 to $200 price will show whether they stay.

Our take: if Wonder Full justifies $50 to $200 a month for coders, Meta has a second business. If it doesn’t, benchmark charts won’t rescue it.

Before you switch anything

  • Run Wonder on your own agent tasks through the API, and track cost per finished task, not just cost per token.
  • Don’t plan around Wonder Full yet, since Meta hasn’t said whether it will ever reach the API.
  • Wait for Meta’s safety report and independent benchmarks before putting Wonder near customer data.

Building on AI agents? Tell us in the comments which model you’d test first, and share this with your tech lead. For more on Meta’s AI push and what it means for Indian startups, keep exploring StartupIndiaX.

FAQs

What is Muse Wonder?

Muse Wonder is Meta’s first frontier-class AI model, released October 12, 2026 by Meta Superintelligence Labs. It’s built for agentic tasks like coding, long-running work and remembering details across conversations, and it succeeds the lighter Muse Spark.

How much does Muse Wonder cost?

Through the Meta Model API, Muse Wonder costs $4 per million input tokens and $20 per million output tokens. In the Muse app, access needs a $20 or $100 monthly plan, according to CNBC.

What’s the difference between Muse Wonder and Muse Wonder Full?

Muse Wonder Full is the stronger variant. It’s limited to Muse Code subscribers on the $50 or $200 plans and isn’t available through the API. Muse Wonder is available through both the API and Muse Code.

Is Muse Wonder available in India?

Meta’s launch announcement doesn’t list regional availability. The Muse app launched in the US first, so check the Meta Model API’s supported regions before planning a product around it.

Who is behind Muse Wonder?

Meta Superintelligence Labs built it, under Meta AI chief Alexandr Wang. The models were trained at Meta’s Prometheus data center in Ohio, and Wang announced them on X on launch day.

Can I trust Meta’s benchmark claims?

Treat them as Meta’s own claims for now. The scores come from Meta’s testing, and Meta says a full safety report is coming. Run your own evaluations on your tasks before committing.

0 comments 31 views
FacebookTwitterLinkedinWhatsapp
Adani and Jabil leaders mark the 50:50 AI hardware joint venture MoU
AI & DeepTechNewsTechnology

Adani and Jabil Just Signed a 50:50 Deal to Build India’s AI Servers

A non-binding MoU, an equal split, and a bet on the hardware that sits inside India’s AI data centres.
by Aalam Rohile
3 min read

SUMMARY

  • Adani Enterprises and Jabil signed a non-binding MoU on October 13, 2026 for a proposed 50:50 JV to make AI data centre hardware in India.
  • The venture would make AI racks, servers, storage, networking, power and cooling equipment, targeting multi-gigawatt capacity for India and for export.
  • Nothing is final: definitive agreements and approvals are pending, and no investment size, location or timeline has been announced. Watch for the binding agreement.

Adani Enterprises and Jabil Inc. have signed a non-binding MoU for a 50:50 joint venture to manufacture and assemble AI and data centre equipment in India. The announcement came on October 13, 2026, four months after the two companies first said they planned to team up.

It’s a big headline with a small asterisk. The idea is large. The deal isn’t done.

Adani Jabil Joint Venture: 50:50 MoU for AI Data Centre Hardware

Here’s the structure. Adani Enterprises, the group’s flagship, and Jabil, the NYSE-listed (JBL) engineering and manufacturing company, would each hold 50% of the new venture. The MoU only sets out the framework, and it builds on the June 2026 announcement of their intent to form a strategic alliance.

Jabil CEO Mike Dastoor says the goal is to bring the company’s rack-level data centre solutions to more hyperscalers, in India and abroad. Jabil reported $36 billion in revenue for fiscal 2026, so this isn’t a small partner.

Adani Wants to Make the Hardware Behind India’s AI Boom, With Jabil as Partner

Until now, Adani’s AI story has been about power, land, logistics and data centres. Gautam Adani says the group now wants to move up the AI value chain, from hosting AI infrastructure to engineering it. His stated aim is to build the “physical backbone of the global AI economy.”

It fits the group’s existing pledge of USD 100 billion for 5 GW of green-energy-powered, AI-ready data centres by 2035. If Adani builds the halls and supplies the electricity, making what goes inside is the logical next piece.

From Racks to Servers: Inside Adani and Jabil’s Plan to Make AI Hardware in India

The product list is wider than “servers” suggests. The JV would make:

  • AI racks, servers, storage and networking equipment
  • Power distribution units, liquid-cooling systems and transformers
  • Installation and commissioning services

The target is multi-gigawatt rack manufacturing capacity. Jabil has been assembling the pieces for a while, buying Hanley Energy Group for power management and Mikros Technologies for liquid cooling.

Read More: Jio Turns Any Old Computer Into a Cloud-Powered PC, Expands JioPC With Plans Starting at Rs 1,000

The Unglamorous Part of AI That Adani Just Bet On: Racks, Servers, Cooling

Nobody tweets about transformers. But a modern AI cluster is as much a power and heat problem as a software one. Dense racks need serious power delivery and liquid cooling.

Whoever can supply all of it as one package saves customers months of juggling vendors. That’s the pitch the companies made in June: an end-to-end, design-to-deployment hardware ecosystem with a single source for infrastructure builders.

India Wants to Stop Importing AI Servers. Adani and Jabil Have a Plan

Demand is the reason. Industry forecasts put India’s data centre capacity at 5 to 8 GW by 2030, driven by AI adoption, cloud expansion and data localisation. The companies also cite more than $50 billion in planned hyperscaler spending across data centres, cloud and AI in India.

Policy is helping. The release points to the Union Budget 2026 tax holiday for data centres until 2047 and the Digital Personal Data Protection Act, 2023. Adani describes the ambition as moving India from importer of digital infrastructure to manufacturer and exporter of AI hardware. The companies size the global opportunity at over $3 trillion across the next seven years.

Why Adani and Jabil’s AI Hardware JV Matters More Than Another Data Centre Announcement

India has no shortage of data centre headlines. Google plans a $15 billion AI hub in Visakhapatnam. In July, HCLTech said it would invest ₹14,257 crore ($1.48 billion) in an Odisha AI data centre with Sarvam AI.

Those stories are about buildings, power and models. This one is about the equipment inside. A country can host the world’s AI compute and still import every rack, which means someone else captures much of the value. A local manufacturing base is meant to change that. It’s an argument, not yet a result.

Read More: India’s First Homegrown Sovereign AI Model, Sarvam AI To Build

Adani and Jabil Eye Made-in-India AI Servers, But the Deal Isn’t Final Yet

Read the fine print. The MoU is non-binding, and the JV remains subject to definitive agreements, approvals and customary conditions. Jabil’s own release warns it may not be able to negotiate and execute a binding agreement.

The announcement also doesn’t give an investment figure, a plant location or a start date. Until it does, treat the multi-gigawatt target as an ambition. This is a developing story.

Adani Jabil MoU Explained: What the 50:50 AI Hardware JV Means for India

In plain terms, two companies agreed to explore building AI data centre hardware together in India. If it goes ahead, India gets a large domestic maker of AI infrastructure, combining Jabil’s 60 years of manufacturing experience with Adani’s energy and logistics reach.

For investors, the thing to watch is the jump from MoU to binding agreement. For the wider ecosystem, it signals that AI hardware manufacturing is now a stated priority for a major Indian group.

What Adani and Jabil’s Hardware JV Means for Indian Founders Building on AI

Founders won’t be buying racks from this JV next month. But if it works, a few things shift. Local supply of servers and cooling could shorten lead times for AI cloud and compute-heavy startups. Power electronics, thermal management and monitoring software are areas where smaller Indian companies could supply or partner.

Sarvam’s tie-up with HCLTech shows the pattern. Big infrastructure players like having Indian AI startups close.

Where this leaves us

Our read: the bet here is on the boring layer of AI, and that’s usually where durable businesses get built. But an MoU is a handshake, not a factory. Three things will tell us if this is real:

  • The definitive agreement, which turns a framework into a company with obligations
  • Disclosure of investment size, plant location and timeline
  • Early signs that local supply of AI hardware reaches startups building compute-heavy products

What do you think, can India build its own AI hardware stack? Tell us in the comments, and share this with a founder who’s watching the AI infrastructure race.

FAQs

What is the Adani Jabil joint venture?

It’s a proposed 50:50 venture between Adani Enterprises and Jabil Inc. to manufacture and assemble AI and data centre equipment in India. A non-binding MoU signed on October 13, 2026 sets out the framework.

What will the JV make?

The venture would make AI racks, servers, storage and networking equipment, plus power distribution units, liquid-cooling systems and transformers. It would also offer installation and commissioning services for data centre customers.

Is the deal final?

No. The MoU is non-binding. The joint venture still needs definitive agreements, applicable approvals and customary conditions to be met before it can be formally set up and begin operations.

Why is India a target for AI hardware manufacturing?

Industry forecasts see India’s data centre capacity reaching 5 to 8 GW by 2030, driven by AI adoption, cloud growth and data localisation. Local manufacturing could serve that demand and support exports.

How does this connect to Adani’s data centre plans?

It aligns with Adani Group’s USD 100 billion commitment to develop 5 GW of green-energy-powered, AI-ready data centres by 2035, adding hardware manufacturing to its infrastructure, energy and logistics presence.

Who leads each company?

Gautam Adani is chairman of the Adani Group, and Mike Dastoor is CEO of Jabil Inc. Both leaders commented on the MoU, which was announced on October 13, 2026.

0 comments 26 views
FacebookTwitterLinkedinWhatsapp
Free Startup Runway Calculator, Startup founder checking cash runway numbers on a laptop at a co-working desk
Startup LearningFounder Toolkit

Free Startup Runway Calculator: You Probably Have Fewer Months of Cash Than You Think

The number in your bank app isn't your runway. Here's how to find the real one, and why Indian founders should start raising with nine months left, not three.
by Aalam Rohile
3 min read

SUMMARY

  • Real runway uses adjusted cash and net burn, including PF, ESI, gratuity and dues you already owe. Your bank balance overstates it.
  • Indian fundraises typically take 4 to 8 months, so start the raise with roughly nine months of runway left, not three.
  • Run your numbers in the free StartupIndiaX Runway Calculator, then stress-test best, base and worst cases before you talk to any investor.

You open your banking app and see ₹1.8 crore. You spend ₹45 lakh a month, collect ₹15 lakh from customers, and the quick maths says you’ve got six months.

Six months sounds survivable. It also isn’t the real number.

Most founders find that out the hard way. Usually it’s month four, and an investor’s “we’re still thinking about it” email is sitting unanswered. The gap between the runway you calculate and the runway you actually have is where startups get into trouble, so let’s close it.

The formula takes ten seconds. The inputs take longer.

Runway is adjusted cash divided by net burn. That’s the whole formula.

Net burn is what you spend each month minus the revenue you actually collect, averaged over the last three months. Adjusted cash is your bank balance after you subtract money that’s already spoken for, like GST and TDS dues or an advance tax instalment.

Here’s an illustrative startup with round numbers (not a real company):

  • Bank balance: ₹180 lakh
  • Monthly spend: ₹45 lakh
  • Revenue collected: ₹15 lakh
  • Net burn: ₹30 lakh

The napkin answer is 180 ÷ 30, or 6 months. Now add what the napkin left out. Employer-side payroll costs that never made it into the burn figure add ₹3 lakh a month, so net burn becomes ₹33 lakh. Committed dues of ₹12 lakh bring adjusted cash down to ₹168 lakh.

168 ÷ 33 gives about 5.1 months. Nearly a month disappeared and nobody spent a rupee.

Use three months of data, not one. A festive-season spike or an unusually quiet month will swing the answer. And recalculate every month, as the eChai Ventures guide suggests, not once a year.

Bar chart showing an example startup's runway falling from 6.0 months to 5.1 months after payroll costs and dues

The costs your burn figure probably misses

Treelife’s burn rate and runway guide points out that many Indian founders count payroll at CTC and miss the employer’s PF and ESI contributions entirely. Those sit on top of salary:

  • PF: the employer pays 12% of basic plus DA, with the wage ceiling at ₹15,000 a month. Add EDLI and admin charges and the all-in cost lands near 13%.
  • ESI: the employer pays 3.25% of gross wages for employees earning up to ₹21,000 a month.
  • Gratuity: accrues at roughly 4.81% of basic.

PF deposits are due by the 15th of the following month. Late payments attract 12% annual interest plus damages. Treelife notes that some Indian startups have had to put in fresh capital to clear PF arrears before a round could close, which is a bad way to spend a cheque you’ve waited months for.

Revenue has the opposite problem. Booked isn’t banked. ₹50 lakh of annual contracts paid quarterly puts ₹12.5 lakh in your account each quarter, and runway runs on collections.

If you sell products, inventory is cash too. Stock bought ahead of a sale season sits on the shelf, not in your account, so a D2C brand can look healthy on paper while the bank balance thins out.

Read More: How to Raise Startup Funding in India (2026)

Your real deadline isn’t zero cash. It’s the end of your raise.

This is the part that should change how you behave. Open’s guide to extending runway says fundraising in India typically takes four to eight months. Take our illustrative 5.1 months. If the raise runs long, the money is gone before the cheque arrives.

One India-focused runway guide on CalcGuru calls this fundable runway: a company with nine months left has roughly four months in which it can still negotiate from strength. After that, you’re pitching with one eye on payroll. Investors notice, too. Walk into a pitch with four months left and they know you need the deal more than they do.

Advice on when to start varies. eChai Ventures’ founder guide says begin the conversation with six or more months left. Other guides say nine to twelve. We’d go with nine, because a raise that takes the long end of the range still closes with a cushion.

Dunzo shows what waiting looks like. It closed a $240 million round in January 2022, with Reliance Retail putting in $200 million, according to Inc42. Entrackr reported, citing sources, that it later needed follow-on money because only a few months of runway remained. Salaries were delayed. A deal that employees were told in July 2024 was in its final stage never closed, Deccan Herald reported, and the app went offline in January 2025.

Dunzo’s burn was far bigger than any seed-stage startup’s, so don’t read it as a forecast for yours. Read it as the price of counting on a round that hasn’t closed.

Timeline showing a 4 to 8 month fundraising window against a startup's zero-cash date, with the start line at nine months of runway

Run three versions, not one

A single runway number is a guess dressed up as a fact. Run three:

  • Base case: trailing three-month burn and collections, exactly as they are.
  • Worst case: collections slip by a month, a key hire starts early, your biggest customer pays late.
  • Best case: the pipeline closes on schedule and revenue grows.

Write down the assumption behind each case. “Collections slip by a month” is something you can test. “Things get worse” isn’t. Once the numbers are on a page, a co-founder or advisor can challenge them before an investor does.

Paul Graham’s idea of “default alive” is a useful test. A startup is default alive if its current growth and burn would reach profitability before the cash runs out. If your best case isn’t default alive, the raise isn’t optional. If even your worst case clears nine months, you can afford to be picky about investors.

If the worst case looks too short, you have levers beyond a round. Cancel unused tools, pause hires, chase receivables. Look at non-dilutive money too, starting with government grant schemes for startups in India.

Read More: Government Grant Schemes for Startups in India

And before you give up equity, see what a round costs you. The Funding Round Dilution Calculator shows how each raise changes your ownership. It sits alongside the rest of the free Founder Toolkit.

Start with the number you’ve been avoiding

Pull three months of bank statements. Work out net burn from collections, not invoices. Add the payroll costs your CTC figure hides, subtract what you already owe, and put the result into the Startup Runway Calculator.

Then count back nine months from your zero-cash date and put that day in your calendar. That’s your start line. Update it monthly, because a runway you only check at year-end tells you about a problem after it has already happened.

  • Subtract committed dues and add employer payroll costs before you divide. Your real runway is often shorter than the napkin number.
  • Fundraising typically takes 4 to 8 months in India, so mark the day you hit nine months of runway as your start line.
  • Run base, worst and best cases in the Startup Runway Calculator, then decide whether to raise, cut costs or do both.

Tried it? Tell us in the comments what your real number turned out to be, and share this with a founder who’s still dividing by the bank balance. For more free tools and funding explainers, keep exploring StartupIndiaX.

FAQs

What is startup runway?

Runway is the number of months your startup can keep operating before cash hits zero, assuming spending and income stay close to today’s levels. It’s your survival window and, in practice, your fundraising deadline.

How do I calculate startup runway?

Divide your adjusted cash balance by your monthly net burn, using a three-month average. Adjusted cash is your bank balance minus dues you’ve already committed. Net burn is total spending minus revenue you’ve actually collected.

How much runway should a startup have?

Most founder guides suggest keeping 12 to 18 months in hand, and a fresh round usually aims to give you 18 to 24. Falling below 12 months means you’re raising from a weaker position with less time to negotiate.

When should I start fundraising?

Start when about nine months of runway remain. Indian rounds typically take four to eight months, and you need a cushion in case the process drags or your first plan falls through.

What’s the difference between gross burn and net burn?

Gross burn is everything you spend each month. Net burn subtracts the revenue you actually collect. Runway uses net burn, so ₹45 lakh of spending against ₹15 lakh collected is a ₹30 lakh net burn.

0 comments 22 views
FacebookTwitterLinkedinWhatsapp
Engineers testing a deeptech hardware prototype in an Indian research lab, as the RDI Fund names its first fund managers
NewsAI & DeepTechFunding

RDI Fund Opens Its Doors to Private VCs: What the First 19 Fund Managers Mean for Founders

The government has named 14 venture funds and 5 research organisations to channel ₹7,131 crore into deeptech. Here's what changes for founders, and what doesn't.
by Aalam Rohile
3 min read

SUMMARY

  • The RDI Fund approved 19 second-level fund managers, 14 AIFs and 5 research organisations, with ₹7,131 crore committed.
  • Managers are expected to match the government’s money with private capital, so founders should prepare for venture-style diligence rather than grant-style paperwork.
  • Projects must sit at Technology Readiness Level 4 or above. Get your IP position, production cost and customer case ready first.

India’s ₹1 lakh crore deeptech fund has finally named the people who’ll decide where the money goes.

On October 12, the Ministry of Science and Technology announced that the Research, Development and Innovation Fund has approved its first cohort of 19 second-level fund managers, with a combined commitment of ₹7,131 crore. The Empowered Group of Secretaries (EGoS) cleared the list on October 8, according to Business Standard and WebIndia.

The fund doesn’t write cheques to startups itself. These managers do. And 14 of the 19 are alternative investment funds, the structure most Indian VC firms use. As of August, only two government bodies, the Technology Development Board (TDB) and BIRAC, were managers, per a Drishti IAS explainer.

Who made the cut

The cohort splits into 14 alternative investment funds and 5 Focused Research Organisations, Inc42 reported. Sources told ET that more than 160 venture capital firms applied. Going by that figure, fewer than 9 in 100 of those applicants got through. That’s our arithmetic, not an official number.

The first call for applications went out in January 2026, and the fund says it will keep onboarding managers in the coming months, so this isn’t the final list, ANI reported.

The matching rule is the real story

Every RDIF second-level fund manager is expected to raise private money equal to what the fund puts in. On paper, that could push the pool past ₹14,000 crore. Treat that as a ceiling. Matching is an expectation until the cheques actually clear.

A manager risking outside investors’ money behaves differently from an agency spending an allocation.

India has run this play before. The Fund of Funds for Startups, set up in 2016 with a ₹10,000 crore corpus at SIDBI, pushed government money into private funds rather than backing startups directly. The ANRF RDI Fund follows the same two-tier logic.

Diagram showing RDI Fund money flowing from ANRF to 19 fund managers and then to deeptech startups, with private capital matching

Why two government bodies weren’t enough

The Cabinet cleared the scheme on July 1, 2025, according to Policy Circle, and it was operationalised on November 3, 2025 over six years.

Early money moved through TDB. On May 13, it signed the first agreements and released the first tranche to five startups: Dhruva Space, Endure Air Systems, ETRNL Energy, Noccarc Robotics and IISTEM Research, Entrepreneur India reported.

Then the pipe clogged. A recent Outlook Business report said TDB had stopped inviting fresh applications, citing administrative reasons. It also said 35 companies had been selected across two rounds, with no letters of intent yet issued to the second batch.

Separately, a written Rajya Sabha reply on July 30 showed seven of the 12 members of TDB’s investment committee had invested in 16 entities selected for funding, The Wire reported. That put conflict-of-interest safeguards under a spotlight.

Read More:

If you build hardware, this was written for you

Deeptech funding in India hit about $2.3 billion in 2025, up 37%, according to a Nasscom-Zinnov study. But AI took roughly 91% of it. If you make batteries, satellites or lab instruments instead of models, you’re fighting over the other 9%.

India’s gross R&D spending was just 0.64% of GDP in 2021, and private companies contributed only 36.4% of it, Policy Circle noted. A government fund for deeptech startups only works if private money shows up too.

What qualifies, and what doesn’t

TRL 4 funding is the entry ticket. Projects must be at Technology Readiness Level 4 or above, meaning the concept has already been validated in a lab. DST’s Jyoti Sharma told Outlook Business that “routine R&D will not be funded,” and that the government can back up to 50% of project cost.

This isn’t a grant, either. The scheme offers low-interest loans, equity or a mix, per Entrepreneur India.

Fund managers also have to judge whether your IP is defensible, whether you can produce at an acceptable cost, and whether a market exists, Policy Circle notes. It also flags a risk: if every failed project triggers an inquiry, managers may drift toward companies that are easier to defend before auditors.

Read More: How to Raise Pre-Seed Funding in India: What Investors Ask in the First 10 Minutes

What we’re watching

This is a developing story. The ₹7,131 crore is an approved commitment, not deployed capital, and nobody has shown that the matching private money has closed.

Three things will tell us whether the model works. How fast the first cheques reach startups. Whether the TDB application pause lifts. And whether later cohorts bring managers with real technical depth, not just fundraising muscle. If they do, India gets a real patient-capital pipeline for science and hardware startups.

Before you pitch any of the 19 RDI Fund fund managers

  • Check your technology readiness level honestly. Below TRL 4, this money isn’t for you yet.
  • Build your IP, production-cost and customer case before any meeting. Managers are expected to test exactly those three.
  • Map other routes now with our Government Scheme Finder and Indian VC & Investor Database at the Founder Toolkit while the new managers ramp up.

Still earlier than TRL 4? Start with our guide to raising startup funding in India.

Building in deeptech and eyeing this route? Tell us in the comments which TRL you’re at, and share this with a founder who needs it.

FAQs

What is the RDI Fund?

The RDI Fund is a ₹1 lakh crore government corpus, approved by the Cabinet in July 2025 and launched in November 2025, meant to give long-term financing to private companies doing research, development and innovation in strategic sectors.

Who are the RDI Fund fund managers?

The first cohort has 19 second-level fund managers approved on October 8: 14 alternative investment funds and 5 focused research organisations. As of August, only the Technology Development Board and BIRAC were acting as fund managers.

How much has the RDI Fund committed to these managers?

The commitment to the first cohort is ₹7,131 crore. Managers are expected to raise matching private capital, so the pool could be larger. Commitments are approvals, though, not money already invested in startups.

Which startups are eligible for RDI Fund money?

Companies working on research-led technology at Technology Readiness Level 4 or above, in areas such as AI, quantum, robotics, space and semiconductors. Officials have said routine R&D won’t be funded; the work has to be innovative.

Is the RDI Fund a grant?

No. Entrepreneur India reports the scheme offers low-interest loans, equity participation or a mix of both, not traditional grants. DST officials say it can cover up to 50 per cent of a project’s cost.

How can a startup apply for RDI Fund money?

The fund doesn’t invest directly; the second-level fund managers evaluate and finance eligible companies. Founders should approach the managers whose focus matches their sector. TDB has reportedly paused fresh applications, so check each manager’s current intake.

0 comments 27 views
FacebookTwitterLinkedinWhatsapp
Creator building a game on a laptop using text prompts, illustrating Google Playground’s no-code game maker
AI & DeepTechTechnology

Want to create games? Google Playground lets you build games without writing a Single Line of Code

Google Labs’ new AI platform turns plain-text prompts into playable browser games, but access is limited to US adults for now.
by Aalam Rohile
3 min read

SUMMARY

  • Google Playground turns plain-text prompts into browser-playable games, with no coding, and lets creators tweak rules, physics and characters through follow-up prompts.
  • It’s live only for US users aged 18+, and creation runs on weekly tokens. Indian creators can’t sign up yet.
  • Meta, Roblox and Unity are racing into AI game creation too. Founders should prototype game ideas on whatever tool reaches them first.

Say you’ve had a game idea for years. Maybe it’s a quiz your friends would love, or a tiny racing game for your brand’s next campaign. The roadblock was always the same: you can’t code, and hiring a developer wasn’t in the budget.

Google just took a swing at that problem. On October 7, Google Labs launched Playground, an experimental platform where you describe a game in plain text and an AI builds something you can play right away in your browser.

How Playground actually works

You don’t open a game engine. You open a chat. Start from a blank canvas, remix a starter prompt, or take guided help to shape the idea.

Next, you pick a genre, say trivia or racing, choose between 2D and 3D, and decide on single-player or multiplayer. Then you describe the mechanics, gameplay and visual style you want.

Google Playground, Illustration showing a text prompt turning into a playable 2D racing game in a browser

The real trick comes after that. Want the physics floatier, the rules stricter or the hero different? You type that too, and you can test the game immediately. You can even upload your own visuals, and the AI turns them into game assets that match the art style.

Google says Playground runs on its own foundation models, including Gemini, Nano Banana and Lyria, plus a custom system refined on games and evaluations built in-house.

Read More: Meta Just Quietly Launched an App Meta Pocket, That Turns Text Prompts Into Playable Games

Sharing, publishing and the Explore gallery

A finished game can stay private, go out as a link, or be published to Playground’s Explore gallery. Published games can carry leaderboards, and some genres support multiplayer. Creators can keep editing after release.

The gallery surfaces games using ratings and play activity. Google says published games go through safety screening, and players can report anything that breaks the rules. With a Play Games profile, you can claim a handle, follow creators and like games.

Here’s the catch for Indian readers

Playground launched for users aged 18 and over in the US only. Google says more countries are coming, but no date for India has surfaced so far. An independent breakdown of Google’s help pages found that no setting or paid plan unlocks it from outside the US.

World map highlighting the US as the only region where Google Playground is available at launch

Playing is free. Making games runs on a weekly token allowance. The free tier gets a limited number, while Google One AI subscribers get higher limits. Google hasn’t published exact figures, so treat any specific number you see elsewhere with caution. The same breakdown says free accounts may wait on a list before they can create.

If you’re eligible, the path is short: sign in at playground.google with your Google account, pick a genre or start blank, describe your game, test it, refine with follow-up prompts, then share or publish.

Read More: College ID Benefits for Indian Students: Free AI Tools You Can Claim Right Now, Before 2027

Everyone wants to be where games get made

Google isn’t alone. Meta has announced AI game creation tools for Horizon, with games potentially recommended across Facebook and Instagram. Roblox has a Build feature that lets creators make games through natural-language prompts. Unity Spark, in public alpha in selected countries, will let several people build a game together and share it on Playground.

Google has played in this space before, from cloud gaming to YouTube Playables, its hub of instantly playable web games. Playground looks like a bet that the next wave of games comes from people who never learned an engine.

What it means for founders and marketers

Our take: the interesting part isn’t the games, it’s the speed. A D2C brand could sketch a quiz or a tap-to-win mini game in an afternoon instead of briefing an agency for weeks. Indian casual hits like Ludo King show how far a simple, familiar format can travel.

There’s a catch worth watching. A reviewer at Droid Life noted that some early catalog games already look obviously AI-made, and warned against a flood of Flappy Bird clones. If everyone can generate a game, the idea becomes the product.

Google for Startups India has been pushing the same message to founders. Its India head has talked about “dismantling traditional coding barriers” for early-stage teams. Playground is that idea applied to games.

Read More: Meta Muse Charm: The Keychain AI Gadget That Could Change How India Uses Apps

What to watch next

Three things matter now: when Google opens Playground beyond the US, what the token limits turn out to be, and whether the Explore gallery fills with original ideas or clones.

Until India gets access, the smart move is to draft your game concept on paper. Write down the genre, the rules, the look and the win condition. When access opens, you can paste it straight in.

Here’s the short version:

  • Playground turns text prompts into playable browser games, with editing and sharing built in.
  • Access is US-only and 18+ for now, so Indian creators must wait for expansion.
  • Meta, Roblox and Unity are moving the same way, so prototype ideas early.

Would you build a game for your brand if access opened in India tomorrow? Tell us in the comments, share this with a founder who keeps saying “I can’t code,” and keep exploring StartupIndiaX for more AI and startup updates.

FAQs

What is Google Playground?

Google Playground is an experimental platform from Google Labs where you describe a game in plain text and an AI builds a playable, browser-based version. You can then tweak it, share it or publish it.

Is Google Playground available in India?

Not yet. At launch it’s open only to users aged 18 and above in the US. Google says more countries are coming, but it hasn’t named India or given a date so far.

Is Google Playground free?

Playing games from the catalog is free. Creating them runs on a weekly token allowance. The free tier gets a limited number of tokens, and Google One AI subscribers get higher limits.

Do I need to know coding to use it?

No. You describe the game in everyday language, pick options like genre, 2D or 3D and single or multiplayer, then refine it through follow-up prompts.

Can I publish and share my game?

Yes. Games can stay private, be shared via a link or be published to the Explore gallery. Published games go through safety screening, and some genres support leaderboards and multiplayer.

Which AI models power Playground?

Google says Playground runs on its own foundation models, including Gemini, Nano Banana and Lyria, plus a custom system refined using games and evaluations built in-house.

0 comments 24 views
FacebookTwitterLinkedinWhatsapp
Indian founder building a startup pitch deck on a laptop with slide notes beside it
Founder ToolkitStartup Learning

How to Build a Startup Pitch Deck in India: 12 Slides to Get Right Before You Email Any Investor

A pitch deck doesn’t close the round. It earns the next meeting. Here’s how to build twelve slides that do exactly that.
by Aalam Rohile
3 min read

SUMMARY

  • Build the deck as 12 slides: opener, problem, solution, market, product, traction, model, competition, team, financials, ask, close. One message per slide.
  • Size your market bottom-up, with customers and annual spend, because investors discount top-down claims like 1% of India.
  • Send a PDF under 8 MB, keep claims reconcilable to your books, and spend your time on traction, not design polish.

Most founders treat the pitch deck like a design project. Investors treat it like a filter.

YourStory’s January 2026 pitch deck playbook says Indian VCs see hundreds of decks a month and spend under four minutes on each. That’s the playbook’s claim, not a measured figure, but the point holds: your deck gets skimmed, not studied.

So the deck has a smaller job than you might think. It doesn’t close a round. It earns the next meeting. Here’s how to build twelve slides that do that, and where Indian investors tend to push back.

Test the story before you open PowerPoint

Here’s a check from the same playbook that’s worth running first. Can you explain your startup in 30 seconds to someone who knows nothing about your industry? If they don’t get the problem and the fix, no slide design will rescue it.

Twelve-slide startup pitch deck structure shown as a storyboard

Once that passes, build the deck as a sequence of questions an investor is already asking in their head. Is the pain real? Is the market big enough? Is there proof? Can this team deliver?

The first three slides: hook, pain, fix

Slide 1, the opener. One sentence and, if you have it, one number. YourStory’s advice is to show a clear outcome instead of a vague ambition. “AI for education” tells an investor nothing. A specific result for a specific number of students does.

Slide 2, the problem. Investors back painkillers, not vitamins. Make the pain feel expensive, frequent and urgent, and back it with verifiable data. Skip problems you’ve never lived with yourself.

Slide 3, the solution. Don’t list features. Show the before and after: what life looks like today, and what changes once someone uses your product.

The market slide is where founders lose credibility

Slide 4. Both YourStory and a CA-led guide from CalcGuru say the same thing: Indian investors discount inflated top-down numbers. The fix is a bottom-up calculation. How many customers exist, what do they pay each year, and what share can you realistically win?

Comparison of top-down and bottom-up market sizing for a pitch deck

A headline figure like the $56 billion baby-care market projection quoted in coverage of Kiddo’s ₹12.5 crore pre-seed round sounds big. But an investor still wants to know how many parents you can reach from your first few dark stores. That’s the number the slide should answer.

Put your assumptions on paper before they go on the slide. The StartupIndiaX TAM/SAM/SOM Calculator is a free way to do that.

Show what works before you show what’s coming

Slide 5, the product. Show what exists today, not what’s on the roadmap. Screenshots of the real thing beat architecture diagrams.

Slide 6, traction. If you have any, this is the slide that matters most. YourStory’s view is that even small numbers beat big claims with no proof: a growing waitlist, early revenue, repeat usage, month-on-month growth. The CalcGuru guide goes a step further and argues investors care more about quality, like retention and revenue concentration, than raw size.

Here’s the uncomfortable part. YourStory’s closing advice is that if traction isn’t there yet, pause the raise and go build it. A mediocre deck with momentum beats a perfect deck with nothing behind it.

The money slides should be boring and believable

Slide 7, the business model. Say exactly how you make money. “We’ll figure out monetisation later” ends the conversation. YourStory suggests a lifetime value of at least three times your customer acquisition cost as a rule of thumb, and says even early estimates beat silence.

Slide 8, competition. Never write “we have no competitors.” It signals you haven’t looked. Name the strong players, say where they’re good, and explain your difference. A comparison chart where you win every row doesn’t help your credibility.

Slide 10, financials. Keep it to an 18 to 24 month operating view with visible assumptions on hiring, pricing and acquisition cost. Going from zero to ₹50 crore in three years with no explanation is the classic hockey stick that costs you trust.

Why this team, and why this amount

Slide 9, the team. Indian investors back people as much as ideas. Show relevant experience and complementary skills. YourStory notes that even a failed earlier startup can add credibility. The CalcGuru guide adds a practical flag: part-time founders tend to get discounted.

Slide 11, the ask. Be specific about how much, what it buys and which milestone it reaches. YourStory’s own example ties ₹2.5 crore at a ₹12 crore pre-money valuation to 18 months of milestones.

Sources split on whether the valuation belongs in the deck at all. CalcGuru suggests stating the amount only and letting pricing come out of the process. We’d say the milestone matters more than either choice. If your raise is at the pre-seed stage, our guide on how to raise pre-seed funding in India covers the cheque sizes and terms behind this slide.

Slide 12, the close. Repeat your core message, your key metric and the ask, and add contact details. If you include a roadmap, keep it to the next 12 to 24 months. Put the cap table, detailed cohort data and any regulatory position in the appendix, not the main arc.

Small details that get decks passed over

  • Format: Send a PDF, not an editable file, under 8 MB. The CalcGuru guide suggests 15 slides at most plus an appendix, while YourStory says most funded decks run 10 to 12.
  • Claims: Every number should reconcile to your books. That’s the standard diligence applies.
  • Cap table: Founders’ majority intact, a clean ESOP pool and DPIIT recognition signal an organised company.
  • Cost: If you’re recognised under Startup India, its portal lists a PitchDeck partner offering one free deck build for recognised startups. Check the current terms first.

For a free structure to start from, the Founder Toolkit at StartupIndiaX includes a Pitch Deck Structure Tool.

Before you hit send

  • Run the 30-second test, then cut any slide that doesn’t answer an investor’s question.
  • Rebuild your market slide bottom-up, with every assumption visible.
  • Make sure your ask names a milestone, and save the PDF under 8 MB.

Which slide gives you the most trouble? Tell us in the comments, or share this with a founder who’s about to start raising. For more tools and funding explainers, explore the StartupIndiaX Founder Toolkit.

FAQs

How many slides should a startup pitch deck have in India?

YourStory says most funded Indian decks run 10 to 12 slides. A CalcGuru guide allows up to 15 plus an appendix. Twelve main slides with the detail pushed into an appendix is a safe middle.

What should a pre-revenue pitch deck focus on?

Team, problem and market carry the most weight when there’s no revenue. Show whatever evidence you have: a waitlist, pilots, or early customer conversations. Investors would rather see honest early signals than invented projections.

Should I put my valuation in the pitch deck?

Sources disagree. YourStory’s example includes a pre-money valuation, while CalcGuru advises stating only the raise amount. Whichever you choose, tie the ask to a specific milestone so investors see what the money unlocks.

What file format should I send to investors?

Send a PDF, not an editable file. A CalcGuru guide suggests keeping it under 8 MB and 15 slides or fewer plus an appendix, so it opens easily on a phone.

Is there a free pitch deck tool in India?

The Startup India portal lists a PitchDeck partner that gives recognised startups one free deck build. StartupIndiaX’s Founder Toolkit also has a free Pitch Deck Structure Tool and a TAM/SAM/SOM Calculator.

What is a good LTV to CAC ratio for a pitch deck?

YourStory suggests a lifetime value of at least three times your customer acquisition cost as a rule of thumb. Even a rough early estimate is better than leaving the business model slide vague.

0 comments 30 views
FacebookTwitterLinkedinWhatsapp
Indian founder pitching a pre-seed funding round to an investor across a café table
Startup LearningFounder Toolkit

How to Raise Pre-Seed Funding in India: What Investors Ask in the First 10 Minutes

The raise takes longer than most founders plan for. Here’s who writes the first cheque, how big it is, and what investors are really weighing.
by Aalam Rohile
3 min read

SUMMARY

  • Expect a 14 to 22 week raise and 20-plus investor conversations. Start your list early and run it in parallel, not one by one.
  • Cheque sizes vary widely, from ₹25 lakh to ₹3 crore depending on the investor. Pick investors by stage and cheque size first.
  • Investors back the team and the problem at this stage. Walk in with proof you can execute, plus clear round terms.

You’ve built a prototype, maybe signed up a few early users, and now you want your first outside cheque. Here’s the part no pitch deck template tells you: the raise itself is a job, and it runs longer than most founders plan for.

A 2026 pre-seed playbook from Eximius Ventures and 1Lattice, as summarised by Backrr, puts a typical Indian pre-seed round at 14 to 22 weeks. The same summary says the median founder converts one term sheet for every 22 to 28 first conversations.

Keep those two numbers in your head while you read the rest.

What “pre-seed” means in India, and why the label is so stretchy

Pre-seed is the first outside money a startup takes, usually before meaningful revenue. Most of it comes from angels, angel syndicates and micro-VCs, and it’s often structured as a convertible note so nobody has to fight over a valuation too early.

Chart comparing pre-seed cheque sizes in India by investor type

The cheque size depends on who you ask. StartupGrantsIndia’s glossary puts typical angel-led pre-seed rounds at ₹5 lakh to ₹25 lakh. One analysis citing Finval puts the typical pre-seed cheque at ₹25 lakh to ₹2 crore, while a Backrr guide gives ₹50 lakh to ₹3 crore for dedicated pre-seed funds.

Then there’s Kiddo. The Delhi NCR baby-care quick-commerce startup announced a ₹12.5 crore pre-seed round led by Campus Fund on September 17, with strategic angels joining. Its founder, Ankit Kawatra, previously built Feeding India, which Zomato acquired in 2019.

That’s a pre-seed round, but not the one most first-time founders will raise. A proven founder, plus a business that needs dark stores and customer acquisition, can stretch the label a long way. Don’t anchor your ask to the biggest headline you saw this month.

Read More: Pre-Money vs Post-Money Valuation: Why the Same ₹10 Crore Costs You Different Equity

Who actually writes the first cheque

You have more options than a decade ago, and that’s both the good news and the problem.

Blume Ventures’ March 2026 micro-VC list tracks roughly 200 sub-$100M funds writing pre-seed to pre-Series A cheques in India. In its earlier 2024 count, Blume put average micro-VC cheques at $200,000 to $300,000. Antler India, which Blume lists separately, writes $500,000 to $1 million cheques.

Dedicated funds keep arriving too. First Cheque closed its second India-focused pre-seed fund at ₹38 crore (about $5 million), with plans to back 50 startups and a focus on first-time founders, according to Business Standard.

Then there’s government money, which founders often misread. The Startup India Seed Fund Scheme can provide up to ₹50 lakh and is routed through incubators, and most schemes need DPIIT recognition first. The ₹10,000 crore Fund of Funds doesn’t invest in startups directly. It puts capital into SEBI-registered funds, which then back you.

If you want to filter by sector, stage and cheque size before you write a single email, the Founder Toolkit at StartupIndiaX has a free VC and investor database built for exactly that.

The five questions behind every first call

No investor hands you a checklist. But the common thread in how pre-seed investors describe their job is blunt: at this stage they’re betting on the team and the problem, not on revenue.

So here’s our read on what’s being weighed in those first ten minutes. It’s an editorial framework, not a survey of investors.

  1. Why you? What makes this team the one to solve it? Past operating experience, a domain you’ve lived in, or a unit you’ve already built.
  2. Why this problem, and why now? A sharp, specific customer pain beats a big market slide every time.
  3. What have you already proved? Even light evidence helps: a working MVP, a pilot, a waitlist that behaves like real demand.
  4. What will this money buy? Investors want to see where the cheque takes you before the next round. Kiddo’s announced plan, covering customer acquisition, dark stores, technology and hiring, is a clean example of a use-of-funds answer.
  5. What are the terms? Round size, instrument, cap and how much of the company you’re selling.

That last one catches first-timers off guard, so it gets its own section.

Read More: Government Schemes for Startups in India 2026: The Guide That Tells You What Not to Apply For

Notes, caps and how much equity you’ll give up

Pre-seed money often arrives as a convertible note. A lower valuation cap means the investor gets more shares for the same cheque, so the cap matters as much as the amount.

One 2026 guide suggests a sensible total of 10% to 20% at pre-seed. Another puts working norms at 8% to 15% for an angel round and 15% to 20% at seed. Treat both as ranges, not rules.

Here’s the catch. Most pre-seed rounds aren’t a single cheque. They’re five or six, signed over months, each at its own cap. Track every one as you go, and run the numbers in the Dilution Calculator before you say yes to the next.

How the raise really runs: weeks, not days

Remember the 14 to 22 weeks? That timeline assumes you’re running a parallel process with 20 or more parties, because the market has grown in participation, not necessarily in conviction.

It also means your runway has to outlast the raise. If you start talking to investors when cash is already tight, you’ll negotiate from the weakest position. Check your numbers in the Runway Calculator first, and begin the process while you still have room to say no.

The stakes are real. The same playbook says fewer than 20% of Indian startups reach Series A within four years. Pre-seed money has to get you to seed-stage proof, not just a few more months of salary.

A few ways first raises stall, based on the guides we reviewed:

  • Raising too little to reach the next milestone
  • Giving up too much equity too early
  • Losing track of who committed what, and at what cap
  • Pitching before there’s an early product to show

Before you send the first email

  • Build a target list of 20-plus investors, filtered by stage, sector and cheque size.
  • Decide your round size, instrument and cap range before the first call, not during it.
  • Plan for a 14 to 22 week process and make sure your cash lasts past it.

Raising a first round right now? Tell us what’s slowing you down in the comments, or share this with the co-founder who keeps asking when the money will land. For more tools and funding explainers, explore the StartupIndiaX Founder Toolkit.

FAQs

What is pre-seed funding in India?

It’s the first outside money a startup raises, usually before meaningful revenue. It often comes from angels, syndicates or micro-VCs, frequently as a convertible note, and is meant to get you to a working product and early users.

How much can you raise at pre-seed in India?

Sources put the typical range anywhere from ₹5 lakh to ₹3 crore, depending on whether the investor is an angel or a fund. Founders with a strong track record can raise more, as Kiddo’s ₹12.5 crore round shows.

How long does a pre-seed round take in India?

One 2026 playbook, as summarised by Backrr, puts it at 14 to 22 weeks. Founders often run conversations with 20 or more investors in parallel, so plan your runway around the full process.

How much equity should you give up at pre-seed?

One 2026 guide suggests 10% to 20% in total across the round. Since pre-seed is often several cheques at different caps, track dilution as you sign. The StartupIndiaX Dilution Calculator can help you model it.

Do you need revenue to raise pre-seed funding?

Not usually. Investors at this stage mainly back the team and the problem. You do need some proof, like an early product or real user interest, to show you can execute.

Can the government fund a pre-seed startup?

Partly. The Startup India Seed Fund Scheme offers up to ₹50 lakh through incubators and generally needs DPIIT recognition. The Fund of Funds invests through registered funds, not directly into startups.

0 comments 35 views
FacebookTwitterLinkedinWhatsapp
Newer Posts
Older Posts