Indian Browsers and Search Engines in 2026: What Actually Works

The Government of India ran a national competition to build an indigenous web browser. It drew 434 teams, ran three rounds over about eighteen months, and Zoho won it. The Press Information Bureau release is unambiguous: "Zoho Corporation emerged as a winner with Team PING- a startup as 1st Runner-Up and Team Ajna a startup as the 2nd Runner-Up... an award of ₹1 crore, ₹75 lakhs, and ₹50 lakhs was given" (PIB, 20 March 2025).

Then very little happened.

Statcounter's July 2026 figures for India put Chrome at 89.19% of browser use, Opera second at 3.76%, Safari third at 3.11%. No Indian browser is a named line item — JioSphere, Ulaa and Epic all sit inside "Other", at around 0.02% combined, despite JioSphere alone having more than 10 million installs. On the search side it is starker still: Google holds 97.85% of Indian search, Bing 1.18%, DuckDuckGo 0.41%. No Indian search engine registers at all.

This article is what we found when we checked each of the four options rather than repeating their marketing. Two of them are in worse shape than any existing write-up says, and one of those findings was significant enough that we verified it two independent ways before publishing. Every figure below was checked directly on 20 August 2026 against the source named next to it.

First, a distinction most articles skip

A browser and a search engine are not the same product, and three of the four things here are browsers.

Installing Ulaa, JioSphere or Epic does not stop you using Google. They are applications for viewing web pages; what they change is the ads you see, the trackers that load, and who receives your telemetry. Unless you also change the default search engine inside them, your queries still go to Google.

Only Qmamu presents itself as an Indian search engine. That turns out to be more complicated than it looks.

The maintenance test

Before any feature comparison, one check matters more than all of them: when was it last updated?

A browser is the largest piece of attack surface on your phone. It renders untrusted code from strangers all day. An unmaintained browser is not a privacy tool, whatever the marketing says — it is an unpatched one. So here is the entire field on that one axis, from each app's own store listing:

App	Package	Android last updated	iOS last updated	Rating	Ratings	Installs
Ulaa	com.zoho.primeum.stable	19 Aug 2026	7 Aug 2026	4.6	33,053	1M+
JioSphere	com.jio.web	14 Aug 2026	23 Jul 2026	4.0	111,498	10M+
Epic Privacy Browser	com.epic.browser	13 Jul 2025	21 May 2020	3.7	12,902	1M+
Qmamu Browser	com.qmamu.browser	18 Mar 2023	— none —	4.6	5,840	100K+

Two of these are actively developed. Two are not. Epic's Android build is thirteen months old and its iOS build is six years and three months old. Qmamu's Android app has not been touched in three years and five months.

That single column reorders the field before anyone reaches a feature list, and it is why the two highest-rated apps in this table are not the two we recommend. Ratings are lifetime averages and go stale. A build date does not.

Ulaa — the one to actually install

Run by: Zoho Corporation, Chennai. Swadeshi Meter score 93.

Ulaa is the IWBDC winner and, on the evidence, the only Indian browser being developed at a serious pace — its Play listing was updated on 19 August 2026, which was yesterday as this was written, and the iOS build shipped on 7 August. Zoho markets a 24-hour security patch policy, and the release cadence is at least consistent with it.

The ownership story is about as clean as this category gets. Zoho is privately held, with founder Sridhar Vembu reported to hold roughly 88%, and no outside investors to answer to — which is precisely why it can afford to ship a browser that makes no money.

What works well

• Genuinely maintained, with the shortest patch cycle claimed by anyone here
• Ad and tracker blocking, a privacy report and fingerprint blocking, on by default
• The only one here evaluated by an independent technical panel (MeitY/C-DAC) on security criteria
• No advertising business anywhere in the parent company, so no structural appetite for your browsing data

Watch out for

• "Privacy-first" is not "collects nothing", and Zoho says so itself. Its EULA, section 6: "Zoho, by default, collects certain details pertaining to your usage of the Browser such as the version of the Browser that you are using, Operating System of your device and the configuration of the device". That is modest, ordinary telemetry — but it is on by default, and any page telling you Ulaa has "zero data collection" has not read Zoho's own agreement.
• It is proprietary, not open source. The same EULA has Zoho owning "all right, title and interest in and to the Browser" and imposes confidentiality over "the Browser's source code". A privacy browser nobody outside the company can audit is a real limitation, and it is the one place Firefox and Brave are structurally ahead.
• It is a Chromium fork — the Play listing bundles Chromium builds — so it inherits Chromium's engine and Google's web-platform decisions
• 1M+ installs against JioSphere's 10M+: a smaller user base means fewer eyes finding bugs

Pricing: Free | Ulaa on SwadeshiApps

JioSphere — the one with actual scale

Run by: Jio Platforms Limited, Mumbai. Swadeshi Meter score 88.

Formerly JioPages, JioSphere is by a wide margin the most-installed Indian browser — 10M+ installs, 111,498 ratings — and it is properly maintained on both platforms. Its pitch is a built-in VPN, ad blocking, and support for "20+ regional Indian languages" with regional news feeds, a genuine differentiator in a category where everyone else ships an English-first product.

Watch out for

• The ownership footnote nobody mentions. Jio Platforms is not wholly Indian-owned. Google invested ₹33,737 crore for a 7.73% stake in 2020 — Reliance's own press release is titled "Google to Invest ₹33,737 Crore in Jio Platforms for a 7.73% Stake" — and Meta holds roughly 10%. So India's leading alternative to Google Chrome is part-owned by Google. That does not make JioSphere a bad browser, but it makes "swadeshi" a more complicated word than the marketing suggests, and it is why its Swadeshi Meter score is 88 rather than Zoho's 93.
• The VPN is ad-gated, and its operator is undisclosed. Jio's own page states: "Based on your choice, you will have to see few short ads." Who runs the exit nodes, and what is logged there, is not published anywhere. A VPN is only as private as whoever operates it.
• The privacy policy is broad. Jio's own in-app policy discloses collection of name, age, gender and contact details, location via GPS and Bluetooth/Wi-Fi signal analysis, device diagnostics, IP address and cookie identifiers — and states that login credentials may be "shared with third parties". Notably that policy names no legal entity at all; it refers only to "we/our/JioSphere App".
• Also a Chromium browser — Jio's own hosted policy path gives it away (.../JioPages/Chromium/Mobile/privacypolicy.html)
• Rated 4.0 against Ulaa's 4.6, on ten times the install base

Pricing: Free | JioSphere on SwadeshiApps

Epic Privacy Browser — still selling a feature it removed a year ago

Run by: Hidden Reflex. Swadeshi Meter score 67 — the lowest here.

Epic is the oldest and best-known name in this group. It is also, on today's evidence, the one we would most strongly tell you not to install — and the reasons come entirely from Epic's own properties.

Watch out for

• Epic discontinued its VPN in July 2025, and is still advertising it today. Epic's own Google Play release notes for the 13 July 2025 update say: "Unfortunately, we are discontinuing the VPN service as it's gotten too expensive and time-consuming to operate, so it's removed from this update". Yet epicbrowser.com/androidfeatures, fetched on 20 August 2026, still says in the present tense: "Free Encrypted Proxy (VPN for the browser)... powered by hundreds of servers in eight countries including the United States, Canada, the United Kingdom, France, Germany, the Netherlands, and Singapore." And the iOS listing is still literally titled "Epic Privacy Browser (w/ VPN)". If the VPN is why you were going to install Epic, it is not there.
• The iOS app was last released on 21 May 2020 — version 1.2, six years and three months ago. For a browser sold on security, that is not a stale build; that is an abandoned one.
• The Android build is thirteen months old, and the website still carries a "© 2025" notice.
• Epic Search is not Epic's own index, by Epic's own admission. Its privacy policy says EpicSearch "sends only your search query to our search API partner to retrieve search results". The partner is never named anywhere on Epic's site.
• It is also the least Indian entry here, which the branding does not suggest. There is a real Indian entity — Hidden Reflex Web Software Pvt Ltd, CIN U72300KA2011PTC060985, registered with the ROC in Bangalore since 2011. But Epic's own legal pages name no jurisdiction at all, and its Terms are "governed by the laws of the United States" with no state specified. Founder Alok Bhardwaj was born in Texas, US-based Graham Holdings has held a stake alongside angel investors with the split undisclosed, and the company also runs operations from Washington DC. Hence the 67.

Pricing: Free | Epic Privacy Browser on SwadeshiApps

Qmamu — Indian company, Google search results

Run by: Qmamu, Ahmedabad. Swadeshi Meter score 93 — and that score is about the company, not the product.

On ownership, Qmamu is the most straightforwardly Indian entry in this article and it is not close. Incorporated in Ahmedabad on 19 January 2021 (CIN U72900GJ2021PTC119605) and launched on Republic Day that year, promoters hold 92.83%, it has taken no external or foreign funding at all, both directors are Indian, and the team is around seven people entirely in India. On ownership criteria alone it is the strongest entry on this page.

The website is live and returns results. But the results are not Qmamu's.

Qmamu's search results are served by Google Programmable Search

We checked this two independent ways, because it is the most important claim in the category and we did not want to get it wrong.

First, a live network capture of qmamu.com/search shows the browser loading cse.google.com/cse.js with the custom search ID cx=2af86af473cd541d4, then calling cse.google.com/cse/element/v1 with the user's query as a parameter. The results rendered on the page are that response. The same call fires on the All, Images, News and Videos tabs.

Second, and verifiable by anyone reading this, Qmamu's own served stylesheet contains rules targeting Google's Custom Search Element classes, scoped to its results container — including #QmamuAll .gsc-control-cse, .gsc-results, .gsc-adBlock and .gsc-search-box. gsc-control-cse is the root class Google's Custom Search Element emits. There is no reason to ship CSS restyling it unless you are rendering one.

The same capture shows the query also going to suggestqueries.google.com for autocomplete, to syndicatedsearch.goog (Google AdSense for Search, which is how Qmamu makes money), to a third-party ad network called Siteplug, and to Google Analytics as a search_term parameter.

Why this matters: Qmamu's own privacy policy states "We don't track you", "No Search Query Tracking", "No IP Logging", and pledges "not to share your personal information or search activity with any third parties, advertisers, or data brokers". Those statements sit alongside an architecture that sends the search term to Google's search API, Google's ad network, a third-party ad network, and Google Analytics.

To be fair about it: building on Google Programmable Search is a perfectly normal way to launch a search product, and many small search sites do exactly this. The problem is not the architecture. It is that the dependency is disclosed nowhere on the site, while the product is marketed on independence and privacy. If your reason for wanting an Indian search engine is to keep your queries away from Google, Qmamu does not currently achieve that.

Also worth knowing

• The Android app has been abandoned in practice — last updated 18 March 2023, three years and five months ago. There is no iOS app at all. The website is maintained (its footer reads "© 2026"); the app is not.
• Qmamu publishes no robots.txt on its own domain, no crawler user-agent, no index-size figure, and no technical documentation of a crawler anywhere.

Pricing: Free | Qmamu on SwadeshiApps

What happened to the browser challenge

Since the IWBDC is the reason anyone expects Indian browsers to exist at all, it is worth reporting what came of it.

The goal was never a new rendering engine. MeitY's own launch release (9 August 2023) describes the aim as "an indigenous web browser with its own trust store with an inbuilt CCA India root certificate" — a certificate-trust and sovereignty goal, not an engine one. The phrase "rendering engine" appears in neither the launch nor the results release. All three winners are Chromium forks, and by the challenge's own terms that was never disqualifying. Anyone expecting an Indian Gecko or WebKit was expecting something the contest did not ask for.

Of the three winners, one shipped. Ulaa is covered above. Team PING (₹75 lakh) still exists at ping-browser.com but has pivoted away from consumers — it now sells itself as "a chromium based browser that replaces legacy VDI solutions to provide secure access to company SAAS on any device". Team Ajna (₹50 lakh), "Bharat Web Navigator", has no Play Store listing, no App Store listing, and no download link we could find.

And the challenge's own website is gone. iwbdc.in returns HTTP 404, with DNS resolving to 127.0.0.1 on registrar parking nameservers.

One caveat worth stating precisely, since everyone reports this differently: the PIB release never uses the word "Ulaa". It names Zoho Corporation, the company. "Ulaa won IWBDC" comes from Zoho's own page and from press coverage, not from the government release. Ulaa also launched publicly before the challenge began, so it was not built for the competition.

So what should you actually do

To stop using Chrome: install Ulaa. It is maintained, independently evaluated, the company behind it has no advertising business, and it shipped an update yesterday. If you browse mainly in an Indian language other than English, JioSphere is the better fit and has the scale — provided the Google and Meta stakes in its parent do not defeat the point for you.

Do not install Epic or the Qmamu app. Epic is thirteen months stale on Android, six years stale on iOS, and still advertising a VPN it removed. Qmamu's app has not been updated since March 2023.

To stop using Google Search: there is no Indian option that does this today. That is the honest answer, and the 97.85% figure is not merely habit. Qmamu is a real Indian company doing a real thing, and we would like it to succeed — but a Google Programmable Search front-end is not independence from Google, and it should not be sold as one.

The uncomfortable conclusion, three years after the browser challenge was announced: India has one very good homegrown browser that almost nobody uses, one large one that is part-owned by the company it is meant to be an alternative to, one that has been quietly abandoned while still taking downloads, and no homegrown search index at all. The first of those is a marketing problem and is fixable. The last one is an index problem, and it is not.

---

All Play Store figures pulled with the India locale (gl=IN) on 20 August 2026 — Google returns different ratings by country, so figures elsewhere may legitimately differ. iOS figures from the iTunes Lookup API, India store. Market-share data: Statcounter Global Stats, India, July 2026. Government sources: PIB releases 2113408 (results) and 1947243 (launch). Swadeshi Meter scores are ours; the methodology is published here.