Maya Research
Who we are
The team
We come from the five billion, and we're obsessed with building for them.
Dheemanth Reddy
Co-founder & CEO
Bharath Kumar
Co-founder & CTO
Founders from
Backed by
The problem
The next five billion hold smartphones they can't use.
They can't shop, bank, or get anything done online. Endless menus, English-only screens, forms, typing, buttons — it's not an interface built for them.
The next five billion
Non-English speakers
India · SE Asia · Africa · LatAm
A mother can't order groceries on Zepto — all English, endless taps.
A shopkeeper gives up the moment a form asks him to type in English.
A grandmother needs her grandson just to pay one bill.
The future
Voice is the next interface and Maya is capturing their gateway to the digital world.
They talk to a phone the way they talk to a person, in their own language.
1B
served by today's internet — English, typed
5B
the next five billion — they talk, not type
1 / 9
The flywheel
Our Maya app has already crossed 34M conversations.
A continuous data flywheel — every new interaction generates proprietary data that makes our models sound native.
People so far have used the product to
- →Talk about their daily life
- →Run their small business — creating posters, videos, and talking through decisions
- →Talk about shopping — finding products to buy
34M
real conversations, compounding every day, in Hindi, Telugu, Bengali and 20+ more languages.
↳ tap to see the data
01
People talk
02
Native data
03
Better models + memory
04
More users
People talk to Maya because we have the best models in the world — they sound native.
We just needed $1M to train the most native model, out-engineering labs with 100x the resources.
Source: Voice Arena ↗
Maya Research
Hear it for yourself
Hear it for yourself
Native vs Alien
Listen to Maya 2: Our most native voice model yet
Our model · raised $1.9M
Maya 2 Native
0:00
Raised $191M
Cartesia Sonic 3.5
0:00
Raised $275M
Sarvam Bulbul v3
0:00
A fraction of the big labs' budgets, and already outpacing them on quality.
Maya Research
The technology
The Architecture
We applied the DeepSeek playbook to native voice.
The whole stack is engineered to be cheap to train and cheap to serve. Three ideas do most of the work.
Architecture
Hybrid attention + Parallel decoding. One checkpoint: block-diffusion + self-speculative, with SSM routing that keeps long turns linear-cost. Quantized, single GPU.
Pre-Training
We take full leverage of frontier text models — A Strong Text based backbone that already reasons across languages — and continually pre-train it on trillions tokens of audio we own at source.
Data Advantage
Our consumer app is a data engine — every conversation feeds proprietary, accented speech back into training. Large teacher models add unlimited synthetic data. No one else has this loop.
The result: The most native speech model, which is at frontier quality and a fraction of the cost
Maya Research
Where we double down
The use cases we double down on
A personal agent that acts on your behalf — by voice and a generative visual UI.
Household decision makers
Pays household bills, orders groceries and medicines, and runs end-to-end online shopping — comparing products, placing orders, tracking deliveries, returns and refunds. Books appointments and travel, and manages everyday banking.
Small business owners & employees
Creates localized promotional content — catalogues, festival posters, sales pitches, presentations and WhatsApp campaigns. Answers customer enquiries, captures orders, follows up on payments, finds suppliers, compares quotations, and helps restock the right products.
Local sales representatives
Plans daily visits and travel routes, localizes sales plans, pitches and creatives, and records meetings — capturing questions, answers and action items. Follows up with leads, shares quotations, captures orders and payments, and manages personalized WhatsApp communication.
Maya Research
Talk to Maya
The same voice in — a very different answer
The speech model that talks to the next 5 billion has local context, is native and full-duplex, and can hold real conversations.
The next 5 billion needs a different architecture — local context, data built for them, done better.
आज का voice AI · turn-based
बोलो · रुको · वो बोले · फिर रुको
माँ
अम्मा
की
दवा
खत्म
प्रतीक्षा…
AI
3
दुकानें
मिलीं।
Maya S2S · full duplex
बोलते हुए सुनती है — बीच में टोको, बदलो
माँ
अम्मा
की
दवा
खत्म…
जल्दी
चाहिए
Maya
ठीक
है,
₹120,
2 घंटे।
Same words in. One returns a list of shops; Maya hears the worry, replies in her language — and just gets it done.
Maya Research
The ask
The opportunity
We have raised $1.9M so far. Now $8M can make Maya the default.
$1.9M
Raised so far — already built
- Models that beat OpenAI's latest at Elo
- The most native speech model, 31 languages
- 3M app downloads · 500K model downloads
→
$8M
What $8M unlocks
- Ship Maya 3, a speech-to-speech (S2S) model: it listens and talks in one model, so conversations feel real, not robotic turn-taking
- Deepen the app — memory & generative UI, grow the user base
- Kickstart the enterprise API motion & monetization
- Hire a lean, world-class research, product & GTM team
To achieve the above, we'll invest heavily in a strong tech, product & sales team.
Where this goes
Intelligence, finally for everyone.
Maya Research
one chat away
1 / 21
The data
What's actually inside 34 million conversations.
How we collect it
Every conversation inside the Maya app — consumer and enterprise — opted-in, real usage, not scripted studio recordings. No paid data-labeling vendors.
What kind of data
Full-duplex audio, transcripts, prosody, accent, and code-switching context — across Hindi, Telugu, Bengali and 20+ more languages.
Quality
Real accented speech from real people, not synthetic TTS loops. Filtered and deduplicated — the exact distribution our models are judged on.
How it helps the models
Feeds continual pre-training and the data flywheel directly — every new conversation makes the next model more native.
Other nuances
Large teacher models add unlimited synthetic augmentation on top of this — but the real accented core is what no one else has.
Organic content on YT
People are already using Maya on their own and making how-to videos about it — see the organic content on YouTube ↗.
The next five billion
Maya is becoming the default voice interface for the next 5 billion.
A world of languages
The next five billion speak 7,159 living languages.
A glimpse of the world's living languages · green = the handful today's AI serves · the rest are the next five billion.
In India alone: 1,369 mother tongues, 121 major languages — 22 with any institutional support.
Quality per dollar
Best model anyone can compete with — on 1/100th the capital.
VoiceArena Hindi score
Leaner · less capital raised →
1086
1072
1058
1044
$10B
$1B
$100M
$10M
The cracked corner
Maya$1.9M
Smallest$8M
ElevenLabs$781M
Cartesia$191M
Sarvam$275M
MiniMax≈$1.77B
xAI≈$42B+
Maya delivers frontier voice quality on a fraction of the capital. Every rival raised more—some by four orders of magnitude.
Quality: VoiceArena Hindi (Bradley-Terry Elo). Total disclosed capital through 4 Jul 2026: Maya $1.9M · Smallest $8M · Cartesia $191M · Sarvam $275M · ElevenLabs $781M · MiniMax ≈$1.77B incl. IPO · xAI ≈$42B+ incl. debt.
Why it doesn't split focus
Enterprise is the same model — turned on, not built.
No new build
Same weights, same training loop that already powers the consumer app. Enterprise is a revenue tap, not a second product — zero diverted engineering.
Demand already exists
500K downloads of the open weights — inbound from conglomerates and startups, already hosted on AWS, Cloudflare, FAL and Baseten. Nothing to cold-start.
Why sequence it
We turn on the paid API once the model matures — so it funds the consumer mission instead of forcing another dilutive raise.
Focus stays consumer
One team, one asset, one focus. Enterprise revenue rides on top of the flywheel — it strengthens the consumer bet, doesn't compete with it.
Open-weight leaderboard
Maya‑1, our open-weight model — released 13 months ago, still top 5.
Source: Artificial Analysis ↗