Skip to content
Featured on Startup Fame

Soket AI raises $4 million to build AI models for finance and defence

Soket AI, an Indian startup building reasoning-focused foundation models, has raised $4 million from Bajaj Finserv and Pennant Technologies. The funding will help scale models for finance, cybersecurity and defence, with Bajaj Finance aiming to turn research into industry tools.

Editor, Lazyfounder

Published 3 min read
Soket AI raises $4 million to build AI models for finance and defence
Image: Inc42 via source

30 SEC SUMMARY

  • Soket AI secured $4 million from Bajaj Finserv and Pennant Technologies to scale its indigenous AI models, targeting finance, cybersecurity and defence.
  • The startup builds foundation models focused on mathematics, coding and reasoning, including its 26-billion-parameter model EKA-1.
  • Funding will split equally between research and enterprise deployments, with Bajaj Finance contributing $2 million alongside Pennant Technologies.
  • Soket AI is part of India’s AI Mission and aims to support Indian languages through an open-source AI stack.
  • Competitors like Sarvam AI and ZenteiQ are also developing multilingual AI models with government backing.

The $4 million deal

Soket AI has closed a $4 million funding round. Bajaj Finserv’s Finserv Intelligence initiative and Pennant Technologies each put in $2 million.

What the money will do

The startup will use the capital to expand its homegrown foundation models. These models specialise in mathematics, coding and logical reasoning—core skills for financial underwriting, fraud detection and defence logistics.

Bajaj Finance Vice Chairman and Managing Director Rajeev Jain said the partnership would combine Soket’s research with Bajaj’s industry know-how to create "practical solutions" for Indian enterprises.

How Soket stacks up

Soket faces competition from other Indian AI labs. ZenteiQ AiTech Innovations recently launched two lightweight multilingual models, Dhīr 2B and Dhīr 5B, for edge devices. Sarvam AI, another government-backed player, rolled out Prism, an open-weight decision model built on its 105B-parameter base.

While Sarvam and ZenteiQ focus on language breadth, Soket is doubling down on reasoning depth—especially for regulated sectors like finance and cybersecurity.

Who is Soket AI?

Founded in 2019 by Abhishek Upperwal, Soket AI is one of the few Indian startups backed by the government’s IndiaAI Mission. Last month it released EKA-1, a 26-billion-parameter model trained for reasoning and maths under Project EKA. The startup is also building Prism, an open-source stack that supports Indian languages.

Part of India’s AI push

The funding fits into New Delhi’s push to develop sovereign AI. India’s AI Mission funds a handful of startups, including Soket and Sarvam, to build models that serve Indian languages and enterprise needs. Both companies tap government grants and collaborate with public-sector labs.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

Bajaj Finance’s $2 million bet signals that large Indian financial institutions are now investing directly in AI research—not just piloting off-the-shelf models. For Soket AI, the cash buys GPU time and talent to refine reasoning-focused models that could outperform generic LLMs in finance and defence.

Key takeaways

  • $4 million funding round co-led by Bajaj Finserv and Pennant Technologies, with each contributing $2 million.
  • Soket AI’s EKA-1 is a 26-billion-parameter model built for reasoning and mathematics, not just language.
  • Bajaj Finance joins the round to fuse its industry expertise with Soket’s research, aiming for tangible financial-services applications.
  • The startup is part of India’s AI Mission and competes with Sarvam AI and ZenteiQ in building sovereign, multilingual AI stacks.

Sources

  1. Inc42 · 2026-11-18
    Soket AI Raises $4 Mn From Bajaj Finserv To Expand Its AI Model Horizon

This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.

About the author

Editor, Lazyfounder

Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.

More stories by Tarun Mottlia

Get the LazyFounder Brief

Startup, funding and AI news in a five-minute read. Join the early-access list.

Lazy Founder - Powered by Blogy.in

Contact us

Have a story tip, correction or partnership idea?

Write to us at tarun.kumar@blogy.in or talk to the founder directly. We read every message.

Contact us