India is the second-largest market for voice-AI startup Wispr Flow by both users and paying subscribers, its founder and CEO Tanay Kothari told ET. Arguing against the perception that Indians are reluctant to pay, Kothari said they spend where they clearly see value.
He explained that this is reflected in Wispr Flow’s strong conversion rates, as users prioritise utility over experimentation. The company, which is building AI-led voice productivity tools, is seeing India play a key role in its growth, both in adoption and monetisation.
“If you are doing something really well, people will pay. Once we understood that, our focus shifted to ensuring users feel they are getting what they paid for. That is what we optimised for,” Kothari said.
“It is interesting that India’s conversion rates, or the percentage of users who pay, are very similar to the US. But the mix is different. In the US, annual and monthly subscriptions are roughly 50-50, whereas in India about 80% of users opt for annual plans. Once you build trust and demonstrate clear value, people are comfortable committing for the long term,” he noted.
Founded in 2021, Wispr Flow has raised about $81 million to date across multiple funding rounds from investors such as Menlo Ventures, Notable Capital, 8VC, and a clutch of angels. It was last valued at around $700 million as of November 2025.
The Silicon Valley-based startup, which was founded by Stanford University graduates Kothari and Sahaj Garg as a brain-computer interface hardware platform, has since pivoted to a voice-first software, shaping its current product strategy. Its fundraising over the past two years came amid growing investor appetite for generative AI, with voice-led productivity tools such as Cartesia and Eleven Labs also raising capital.
Kothari, however, said that venture capital firms in the US were stepping back from investing in large swathes of AI software startups‚ marking what could be the “beginning of the end” of the current bubble.
He said that capital is increasingly drying up for companies with weak unit economics, poor retention, low recurring revenue, and fault lines that were previously masked by abundant funding. “If you have built primarily on venture funding and need the next round to survive because your unit economics are deeply negative, then you are in serious trouble. That is what you are starting to see with a lot of companies.”
“The bubble ending is not a bad thing,” he said. “It just means the companies actually building something valuable will stand out.”
India a voice-first market
India is becoming a major growth area for Silicon Valley voice AI startups. In order to meet the demand for real-time, multilingual speech AI, US-based Cartesia is investing $2.5 million to establish operations in Bengaluru. ElevenLabs, one of the biggest voice AI companies in India, is targeting enterprise adoption, as reported by ET.
Kothari said the company’s India strategy is centred on targeting early adopters within the country’s tech ecosystem before expanding to a wider audience.
“We’re starting with users who are most similar to our original base in San Francisco, primarily the tech community in Bengaluru. Once adoption picks up there, it naturally influences the rest of the market, as these users tend to be aspirational for a broader segment.”
The ambition, he said, is to make voice the default interface across Indian workplaces.
“Our goal is to get into every company in India and make voice the default way people work. The way we do that is by targeting well-known companies, landing those accounts, and then expanding from there,” he added.
The startup claims it has already onboarded large technology companies in the US and begun adding new enterprise clients at a rapid pace, including Fortune 500 firms, with several more in advanced stages of discussions.
The defensibility question
Addressing concerns around whether larger AI firms can replicate its product, Kothari said defensibility comes down to focus and execution speed rather than just capital.
Companies like OpenAI and Anthropic are “fighting a hundred battles” across models, AGI, and their enterprise customers, which dilutes attention on voice, he said, adding that even with resources, “you can only do so many things at a given point in time.”
“They are focused on enterprise customers who want fine-tuned LLMs, so most efforts are directed there. If you try OpenAI as a casual desktop app, it feels like something built by a Y Combinator startup, given the number of bugs. That reflects where their focus lies, because if they wanted to make it great, they could,” he said.
“We are building our company with a very different approach. We are solving an extremely specific problem, with a deep understanding of both the use case and the users. That allows us to move quickly and capture as much market share as possible, with a strong emphasis on growth and brand.”
Kothari argued that building high-quality voice models takes time, and expects competitors will take “six to nine months” to catch up with where Wispr is today.
The startup is focussing on deep problems, faster execution, and rapid market capture. Kothari says the real risk is not the competition catching up, but whether the company can “move fast enough” to stay ahead.
He explained that this is reflected in Wispr Flow’s strong conversion rates, as users prioritise utility over experimentation. The company, which is building AI-led voice productivity tools, is seeing India play a key role in its growth, both in adoption and monetisation.
“If you are doing something really well, people will pay. Once we understood that, our focus shifted to ensuring users feel they are getting what they paid for. That is what we optimised for,” Kothari said.
“It is interesting that India’s conversion rates, or the percentage of users who pay, are very similar to the US. But the mix is different. In the US, annual and monthly subscriptions are roughly 50-50, whereas in India about 80% of users opt for annual plans. Once you build trust and demonstrate clear value, people are comfortable committing for the long term,” he noted.
Founded in 2021, Wispr Flow has raised about $81 million to date across multiple funding rounds from investors such as Menlo Ventures, Notable Capital, 8VC, and a clutch of angels. It was last valued at around $700 million as of November 2025.
The Silicon Valley-based startup, which was founded by Stanford University graduates Kothari and Sahaj Garg as a brain-computer interface hardware platform, has since pivoted to a voice-first software, shaping its current product strategy. Its fundraising over the past two years came amid growing investor appetite for generative AI, with voice-led productivity tools such as Cartesia and Eleven Labs also raising capital.
Kothari, however, said that venture capital firms in the US were stepping back from investing in large swathes of AI software startups‚ marking what could be the “beginning of the end” of the current bubble.
He said that capital is increasingly drying up for companies with weak unit economics, poor retention, low recurring revenue, and fault lines that were previously masked by abundant funding. “If you have built primarily on venture funding and need the next round to survive because your unit economics are deeply negative, then you are in serious trouble. That is what you are starting to see with a lot of companies.”
“The bubble ending is not a bad thing,” he said. “It just means the companies actually building something valuable will stand out.”
India a voice-first market
India is becoming a major growth area for Silicon Valley voice AI startups. In order to meet the demand for real-time, multilingual speech AI, US-based Cartesia is investing $2.5 million to establish operations in Bengaluru. ElevenLabs, one of the biggest voice AI companies in India, is targeting enterprise adoption, as reported by ET.
Kothari said the company’s India strategy is centred on targeting early adopters within the country’s tech ecosystem before expanding to a wider audience.
“We’re starting with users who are most similar to our original base in San Francisco, primarily the tech community in Bengaluru. Once adoption picks up there, it naturally influences the rest of the market, as these users tend to be aspirational for a broader segment.”
The ambition, he said, is to make voice the default interface across Indian workplaces.
“Our goal is to get into every company in India and make voice the default way people work. The way we do that is by targeting well-known companies, landing those accounts, and then expanding from there,” he added.
The startup claims it has already onboarded large technology companies in the US and begun adding new enterprise clients at a rapid pace, including Fortune 500 firms, with several more in advanced stages of discussions.
The defensibility question
Addressing concerns around whether larger AI firms can replicate its product, Kothari said defensibility comes down to focus and execution speed rather than just capital.
Companies like OpenAI and Anthropic are “fighting a hundred battles” across models, AGI, and their enterprise customers, which dilutes attention on voice, he said, adding that even with resources, “you can only do so many things at a given point in time.”
“They are focused on enterprise customers who want fine-tuned LLMs, so most efforts are directed there. If you try OpenAI as a casual desktop app, it feels like something built by a Y Combinator startup, given the number of bugs. That reflects where their focus lies, because if they wanted to make it great, they could,” he said.
“We are building our company with a very different approach. We are solving an extremely specific problem, with a deep understanding of both the use case and the users. That allows us to move quickly and capture as much market share as possible, with a strong emphasis on growth and brand.”
Kothari argued that building high-quality voice models takes time, and expects competitors will take “six to nine months” to catch up with where Wispr is today.
The startup is focussing on deep problems, faster execution, and rapid market capture. Kothari says the real risk is not the competition catching up, but whether the company can “move fast enough” to stay ahead.
