Startup Deep Dive : LambdaTest — the testing platform that renamed itself after the AI that could replace it

LambdaTest spent eight years selling access to other people’s browsers and phones so developers could check whether their websites broke. Then, in January 2026, it renamed itself TestMu AI and said AI agents already generate roughly half its revenue, up from zero when the product shipped in August 2024. The company that built its business on human testers clicking through 3,000 browser-device combinations is now betting its future on software that needs no tester at all.

That bet sits on top of a business that is still small by the standards of its main rival. LambdaTest’s India entity reported ₹177 crore (about $18.4 million, at $1 ≈ ₹96.0 as of 18 September 2026, Trading Economics) in revenue for the year ended March 2024, against BrowserStack’s disclosed $225 million in 2025, as per Wikipedia’s company data citing BrowserStack’s own figures. LambdaTest has raised $108 million across six rounds since 2018; BrowserStack raised $250 million and was valued at $4 billion in 2021. The smaller company is the one that shipped an agentic testing product first.

Quick facts

What they do

LambdaTest sells developers and QA teams a cloud full of browsers, operating systems and real mobile devices so they can check that a website or app works the same way on a fifteen-year-old Android phone as it does on the tester’s own laptop. Under the hood it is infrastructure: more than 3,000 browser and device combinations, spun up on demand, that a company would otherwise have to buy, rack and maintain itself as a physical “device lab”. On top of that infrastructure it now sells software that writes and runs the tests for you — HyperExecute, a faster test-execution grid, and KaneAI, an AI agent launched in August 2024 that lets a tester describe a test case in plain English and have it generated, executed and maintained automatically. The customer base spans small product teams paying $15–29 a month up to enterprises with custom contracts; the company said in December 2024 that it served more than 15,000 paying customers and over 2.3 million registered developers and testers across 130-plus countries.

The origin

Asad Khan did not stumble into testing. He had already spent close to a decade in it: as a lead engineer doing QA and business analysis at GlobalLogic, and then as co-founder of 360Logica, an independent software-testing services firm he built up over several years before it was acquired by Saksoft — Saksoft picked up a majority 51% stake in January 2015 and completed the acquisition in June 2016, according to company-history accounts compiled by The Brand Hopper. Running a testing services shop for years exposed him to the same complaint from client after client: there was no single, affordable place to run both manual, interactive tests and automated Selenium test suites across many browsers at once. Existing options were either expensive enterprise tools or a patchwork of self-hosted virtual machines. In early 2017 Khan assembled a small team, including Jay Singh, to build that missing layer as a cloud service rather than a testing services contract. Singh brought the go-to-market side: more than 15 years in sales and business development, including roles at Harman International and LeadSquared, and, in LambdaTest’s earliest months, he reportedly made around a hundred cold calls himself to land the first customers, according to the company’s own retelling of its history.

The struggle years

The first setback was technical and came before the company had a single paying customer. According to accounts of LambdaTest’s early build, the team had to overhaul its architecture midway through development in 2017 after early versions failed to meet basic performance expectations — running dozens of virtual browser sessions reliably is a different engineering problem than running a handful, and the first attempt did not hold up. The rebuild pushed the first public version of the product to November 2017, months later than planned.

The second, longer struggle was commercial rather than technical: finding product-market fit. Khan has said in an interview with Blume Ventures that the team deliberately launched its least ambitious product first — plain cross-browser testing — rather than its more sophisticated automation tooling, reasoning that a genuinely novel product is always slow to win trust from unfamiliar buyers. That entry product got LambdaTest to about 100 paying customers, but at an average contract value of only around $180 a year — not enough to build a durable business on. Real traction did not arrive until the company pushed an adjacent, higher-priced automation product, priced at roughly $1,000 to $1,200 a year per license, and watched a handful of early customers pay that much without hesitation. Khan has described that shift — from a $180-a-year commodity product to a small group of customers willingly paying five to seven times as much for automation — as the moment the company actually found product-market fit, some two to three years after founding.

The turning point

The clearest before-and-after in LambdaTest’s history is that pivot from cross-browser testing as the flagship offering to automation as the flagship offering. Before it: roughly 100 customers on the entry-level product at about $180 in annual contract value each — call it a business generating perhaps $18,000 a year from that cohort, nowhere near enough to support a venture-scale company. After it: the first five to six customers of the automation product alone were paying $5,000 to $25,000 a year each, with per-license pricing around $1,000 to $1,200, according to Khan’s own account of the period. It was a change in what the product did — from letting a human click through browsers to letting a team run large automated test suites in the cloud — and it reset the unit economics of the whole company. Everything LambdaTest built afterwards, including the Series A round that closed in December 2020, sits downstream of that repricing.

The money behind it

LambdaTest has raised $108 million in total across roughly six rounds since 2018, according to Entrackr’s reporting on its December 2024 round and corroborating funding-tracker data. The shape of that capital: an early seed and pre-Series A from Leo Capital and Blume Ventures, a Series A of $6 million in December 2020 from Sequoia Capital India’s Surge programme, a $16 million Series B in June 2021 led by Sequoia Capital India (now Peak XV Partners) with participation from Telstra Ventures and Wamda Capital, a $45 million Series C in March 2022 led by Premji Invest, and a $38 million Series D in December 2024 led by Avataar Ventures with Qualcomm Ventures joining as a new investor, per PRNewswire’s and TechCrunch’s reporting on those rounds. Two backers changed what the company could do rather than just how much cash it had: Peak XV’s Series B arrived as the company was proving automation demand and needed to fund international expansion, and Premji Invest’s Series C — its largest round to date — funded the buildout of HyperExecute, the orchestration engine that underpins its current speed claims. The Series D, its first new capital in about 30 months per Entrackr, was earmarked specifically to scale KaneAI and its broader AI push, and brought in Qualcomm Ventures, a strategic investor with its own interest in on-device and AI workloads. The round valued the company at a reported $400 million, per aggregator data from PitchBook and Tracxn; Forge Global’s separate secondary-market pricing puts it slightly lower, near $392 million. Inc42 classifies LambdaTest as a “soonicorn” rather than a confirmed unicorn, and notes the company itself has not published an official valuation figure — so every number here is a third-party estimate, not a company disclosure.

How it makes money

The business is a straightforward SaaS subscription sold two ways. Self-serve customers pick a tier — plans start around $15–29 a month for basic live and automated testing — priced mainly on how many tests you can run in parallel and how many browser or device minutes you use, with a free tier offering roughly 200 minutes to pull people in before they pay. Enterprise customers negotiate custom contracts that add single sign-on, on-premise or private-cloud deployment, and dedicated support. On top of both tiers sit newer, higher-margin add-ons: HyperExecute, which the company says can cut test-suite run time by up to 70% compared with traditional cloud grids, and KaneAI, priced as part of its AI-native tier, which the company claims can cut manual effort in writing and maintaining test scripts by 40–70%. The part people consistently get wrong is treating LambdaTest’s India-entity financial filings — the ₹177 crore FY24 revenue figure that shows up in Indian business press — as the company’s global revenue. LambdaTest India Private Limited is the Noida-based engineering and delivery arm of a company that is legally headquartered and predominantly sold in the United States; its India filings capture the local entity’s books, not global bookings, which is why third-party analysts estimate LambdaTest’s worldwide revenue at closer to $89–120 million for 2023–2024 — a company-unconfirmed figure roughly five to six times larger than what shows up in the Indian regulatory filings alone. Where the actual margin sits: infrastructure costs (cloud compute, real-device hardware, bandwidth) are the primary variable cost against subscription revenue, and the newer AI features carry inference costs that are less well understood publicly than the older per-browser-session economics.

The numbers

Figures below are for LambdaTest India Private Limited, the company’s Indian corporate entity, drawn from its Ministry of Corporate Affairs filings as reported by Indian business outlets. They are a partial, India-only view of a global business; a full global income statement is not public.

Revenue at the India entity roughly doubled from FY22 to FY23 and grew a further 54.2% into FY24, according to Inc42’s and YourStory’s reporting on the FY24 filing. Profit grew even faster than revenue in that last year — net profit was up about 62% against a 54% revenue increase, which points to the entity’s cost base (largely engineering headcount and infrastructure) growing more slowly than its top line through FY24.

Where the money comes from

LambdaTest does not publish a geography or customer-segment revenue split, so the clearest documented split is by product line, and it is the more interesting one. Until August 2024, essentially all of LambdaTest’s revenue came from infrastructure-style products — browser access, device access, and test orchestration through HyperExecute. Eighteen months after KaneAI’s launch, the company said at its January 2026 rebrand that AI-native products already accounted for roughly half of total revenue, with the other half still coming from the older cross-browser and real-device cloud business. That is the surprise: a company that spent seven years selling access to other people’s hardware now gets about as much revenue from software that writes tests as it does from the infrastructure that runs them. Organisationally, the two halves still share the same customer base — the same 15,000-plus accounts that bought browser-testing minutes are increasingly also buying KaneAI seats — rather than the AI product opening up an entirely new market.

The risks

First, commoditisation from free tooling. Playwright, the open-source browser-automation framework built and maintained by Microsoft, has become the default choice for many teams writing new browser tests, and it costs nothing to license — competing directly with the value LambdaTest’s core cross-browser product has historically charged for. Selenium and Cypress apply similar pressure. LambdaTest’s own business model depends on teams still choosing to pay for hosted infrastructure and orchestration around these free frameworks rather than running them in-house, and that willingness to pay is the single biggest assumption behind roughly half its historical revenue base.

Second, competitive scale. BrowserStack, founded in 2011 by Ritesh Arora and Nakul Aggarwal, has raised $250 million against LambdaTest’s $108 million, was valued at $4 billion in its 2021 Series B, and reported $225 million in 2025 revenue with more than 1,200 employees, per Wikipedia’s company profile citing the company’s own disclosures. That is a materially larger balance sheet and salesforce to compete against for the same enterprise buyers, particularly as both companies now sell overlapping AI testing features.

Third, the AI pivot could cannibalise the company’s own core business faster than expected, or slower than promised. If AI coding agents increasingly write and self-verify their own code, the market for a separate, dedicated QA testing layer could shrink industry-wide, a risk LambdaTest is explicitly trying to get ahead of by rebranding around AI agents rather than around browsers. But that same rebrand carries execution risk: KaneAI has to keep improving quickly enough to justify being roughly half of the company’s revenue base within eighteen months of launch, and a slowdown in that product’s adoption would hit both halves of the business at once, given the two customer bases substantially overlap.

The takeaway

LambdaTest’s most useful lesson is not about AI at all — it is about repricing before you pivot the product. The company did not find product-market fit by inventing a new feature; it found it by launching a more sophisticated version of what it already did and discovering customers would pay five to seven times more for it. The AI shift two decades later follows the same pattern: KaneAI is not a different business, it is a higher-value way of doing the same job — writing and running tests — for a customer base that already existed. Founders chasing a pivot often look for a new market; LambdaTest’s history suggests looking first at whether the existing market would pay more for a better version of what you already sell.

Frequently asked questions

Is LambdaTest the same company as TestMu AI?

Yes. LambdaTest rebranded to TestMu AI on 12 January 2026, according to the company’s own announcement and third-party coverage of the change. The underlying products, pricing, APIs and the lambdatest.com domain continue to function under the new name.

Who founded LambdaTest and when?

Asad Khan, Jay Singh and Mayank Bhola founded the company in early 2017; its first product version shipped in November 2017. Khan had previously co-founded testing services firm 360Logica, which was acquired by Saksoft between January 2015 and June 2016.

How much funding has LambdaTest raised, and what is it worth?

The company has raised $108 million across roughly six rounds since 2018, most recently a $38 million Series D in December 2024 led by Avataar Ventures with Qualcomm Ventures. That round is reported to value the company at approximately $400 million, per PitchBook and Tracxn aggregation, though Inc42 notes LambdaTest has not officially disclosed a valuation and classifies it as a “soonicorn.”

Is LambdaTest profitable?

Its India entity, LambdaTest India Private Limited, reported a net profit of ₹18.04 crore in FY24 (year ended March 2024), up from ₹11.16 crore in FY23, according to filings reported by Inc42 and YourStory. That covers the India entity only; the company has not disclosed global profitability.

How is LambdaTest different from BrowserStack?

Both sell cloud-based cross-browser and device testing plus, more recently, AI-driven testing tools. BrowserStack is older (founded 2011), better capitalised ($250 million raised, a $4 billion 2021 valuation, and $225 million in reported 2025 revenue per Wikipedia), and generally positioned as the higher-priced, enterprise-trust option; comparison sites report LambdaTest pricing plans at roughly 20–40% less than BrowserStack’s equivalent tiers for comparable session volumes.

Sources

Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).

• TechCrunch, “Software testing platform LambdaTest secures $38 million in AI push,” December 2024
• Entrackr, “LambdaTest raises $38 Mn to scale QA platform KaneAI,” December 2024
• PRNewswire, “LambdaTest closes $45 million in a venture round led by Premji Invest,” March 2022
• GlobeNewswire / Nasdaq press release, “LambdaTest rallies global investors raising $16m in series B funding,” June 2021
• Inc42, LambdaTest company and financials profile, accessed September 2026
• YourStory, “LambdaTest India net profit soars 62% to Rs 18 Cr in FY24,” November 2024
• Blume Ventures, “PMF Convo #6 — Asad Khan, Co-Founder & CEO at LambdaTest,” founder interview
• The Brand Hopper, “LambdaTest – Founders, Business Model, Funding & Competitors,” accessed September 2026
• Bug0, “TestMu AI Formerly LambdaTest: The 2026 Rebrand Explained,” January 2026
• PitchBook, LambdaTest company profile, accessed September 2026
• Tracxn, LambdaTest funding and company profile, accessed September 2026
• Forge Global, LambdaTest pre-IPO stock and valuation page, accessed September 2026
• Wikipedia, BrowserStack company profile, accessed September 2026

Found an error? Write to us and we’ll correct it in the open, dated, on the piece.