Kruti AI: Features, Working Model, and Comparison with ChatGPT

What Is Kruti AI?

Kruti is an agentic AI assistant developed by Ola Krutrim, officially released on 12 June 2025 as an upgrade to passive chatbots, with support for text and voice in 13 Indian languages. It was built to replace the company’s earlier Krutrim chatbot, which launched in 2023 as a proof of concept. Where the predecessor could only answer questions, Kruti executes tasks in the real world, from booking rides to paying bills, all via simple text or voice prompts.

Founder Bhavish Aggarwal described Kruti as “the first real step towards the future of AI where technology doesn’t just talk back, but actually helps you get things done,” adding that the system was built to work the way Indians live, multilingual, mobile-first, and intuitive.

The launch positions Krutrim directly against global AI players like OpenAI and Google, as well as domestic competitors including Sarvam AI and CoRover.ai, in what is rapidly becoming one of the world’s most contested AI markets. To understand how Kruti fits within the broader landscape of conversational AI options available to Indian users and businesses, it helps to first examine what makes it structurally different from everything that came before it.

Key Features of Kruti AI

1. Agentic Task Execution

The defining feature that separates Kruti from conventional chatbots is its ability to act, not just respond. As an agentic AI, Kruti can reason, plan, and execute multi-step tasks to fulfill a user’s request. A command like “Book me a cab to the airport like last Monday” is not just answered, it is actioned. By holding a stacked memory of past actions, contexts, and preferences, Kruti acts more like a digital co-worker than a virtual assistant.

This is precisely the direction the global industry is heading. Businesses exploring how AI chatbots can be deployed to handle real customer service workflows , not just answer FAQs, will find Kruti’s agentic architecture a useful reference point for what capable AI actually looks like in practice.

2. Multilingual Support Across 13 Indian Languages

Language coverage is arguably Kruti’s most strategically significant feature. Kruti supports over 10 Indian languages at launch: Hindi, Tamil, Telugu, Kannada, Malayalam, Bengali, Marathi, Gujarati, Odia, and Punjabi, with the full launch covering 13 languages and a roadmap toward 22. Its speech recognition is built to identify regional Indian languages, dialects, and accents, a capability that global assistants, trained overwhelmingly on English and Western languages, have consistently struggled to replicate with cultural authenticity.

This matters at scale. Only 1% of data on the web is in Indian languages, despite the Indian population representing roughly 20% of the global population, meaning that building language models for India requires deliberately curating and generating local training data, not simply scraping the existing web.

For businesses operating in the subcontinent, this multilingual capability directly addresses one of the most persistent gaps in live chat and support tools available in the Indian market — the inability to serve customers naturally in their preferred regional language.

3. Multimodal Input and Adaptive Output

Kruti processes text, voice, images, and files, making it versatile for Indian users who prefer WhatsApp-style communication. On the output side, Kruti formats responses for clarity using summaries, tables, and stories, adapting to the complexity and context of the request rather than defaulting to a single response format. It also features read-aloud capabilities, making it more inclusive and accessible for users who may prefer audio responses or have lower literacy in text-based interaction.

4. Free Premium Capabilities

Krutrim’s decision to offer advanced features such as image generation, research assistance, and read-aloud capabilities at no cost reflects a deliberate approach to scaling AI in a price-sensitive market. Where global competitors typically put image generation and deep research behind paid subscription tiers, Kruti makes these available to all users at launch, a calculated move to drive mass adoption across India’s Tier-2 and Tier-3 cities.

This approach mirrors a broader shift in the market, where AI-powered chatbot solutions that offer genuine functionality at no cost are increasingly setting the baseline expectation for what users consider acceptable before they’ll pay for anything.

5. Persistent Memory and Personalization

Over time, Kruti learns user behaviour, such as commonly visited locations or preferred food choices, to further personalize interactions. This persistent memory layer is what elevates it beyond a voice-activated command interface. The system builds a user model across sessions, which progressively improves the relevance and speed of task execution without requiring the user to repeat context.

6. Developer SDK and Third-Party Integration

Kruti also includes a fully embeddable software development kit (SDK), allowing developers to integrate LLM orchestration, memory handling, and tool execution with minimal code. Kruti connects to company databases and APIs via the Model Context Protocol and presents responses as summaries, tables, or narratives adapted to user behaviour. The system supports payments via credit/debit cards and UPI. Initially, integrations are limited to the Ola ecosystem, with planned expansions to services like Blinkit, Swiggy, and Uber.

For development teams evaluating whether to build or buy, understanding what professional chatbot development and integration actually involves is a useful exercise before deciding whether an SDK-based approach like Kruti’s fits their technical roadmap.

How Kruti AI Works: The Technical Model

The backend technology combines several open-source large language models with Ola’s proprietary Krutrim V2 model, which has 12 billion parameters. This hybrid architecture, mixing open-source efficiency with a proprietary model fine-tuned on Indian data, keeps inference costs low while maintaining performance appropriate for the Indian market.

Kruti uses a combination of agent-to-agent protocols and a Model Context Protocol (MCP) to connect with third-party APIs and internal databases. This allows it to perform complex actions like generating AI images, summarizing documents, or initiating payments with minimal input.

Optimizing for the hardware reality of the Indian market is a core design principle. The system was developed to work primarily on smartphones, addressing the Indian market’s specific needs, including language diversity and potential bandwidth constraints. It is optimized for edge computing, allowing deployment on mobile devices, IoT devices, and smart kiosks, with a scalable backend designed to support millions of concurrent users. Data is hosted on Indian servers, ensuring data privacy and regulatory compliance.

This last point is increasingly important. As governments tighten oversight of AI systems, businesses and platforms operating in regulated markets need to understand the evolving compliance and legal landscape surrounding AI chatbot deployment. Kruti’s India-hosted infrastructure gives it a structural compliance advantage that US-based platforms cannot easily replicate.

Krutrim is also investing in homegrown AI infrastructure, including custom chips called Bodhi and Sarv, set to debut by 2026, signaling ambition to own the full AI stack, from silicon to application layer.

Kruti AI vs. ChatGPT: A Direct Comparison

Both Kruti and ChatGPT are AI systems that handle natural language queries, but they are built for fundamentally different purposes and geographies.

Dimension Kruti AI ChatGPT (GPT-4o) Primary purpose Agentic task execution Conversational AI & reasoning Language coverage 13 Indian languages + English 50+ languages, primarily English-optimized Task automation Books cabs, pays bills, orders food Limited native action capabilities Cultural context Deep Indian cultural training Primarily Western cultural context Platform focus Android-first, mobile-optimized Web, desktop, mobile Image generation Free, included Requires ChatGPT Plus ($20/month) Memory Persistent, cross-session Available in paid tier Underlying model Krutrim V2 + open-source hybrid GPT-4o (OpenAI proprietary) Pricing Core features free Free tier + Plus/Pro tiers Data jurisdiction Indian servers US-based infrastructure

The most meaningful distinction is architectural intent. ChatGPT is a highly capable general-purpose reasoning engine, it is built to think, write, code, and analyze. It does not natively book a cab or initiate a UPI payment. Kruti is built around action. Its value proposition is not the quality of its prose but the completion of real-world tasks through a conversational interface.

When evaluating tools like these, it’s also worth understanding what AI chat tools are genuinely available for free versus what requires a paid subscription, because the gap between Kruti’s free offering and ChatGPT Plus’s pricing is a significant competitive differentiator in price-sensitive markets.

For a software developer in Bengaluru who needs to debug code, write documentation, or analyze a dataset, ChatGPT remains the stronger tool today. For a small business owner in Lucknow who wants to book logistics, check orders, and communicate in Hindi without switching between five apps, Kruti is purpose-built for that workflow in a way ChatGPT is not.

The comparison also extends to data sovereignty, a concern that is increasingly material in enterprise and government AI decisions. Kruti’s India-hosted infrastructure aligns with data localization requirements and Aatmanirbhar Bharat objectives in ways that US-hosted alternatives structurally cannot.

Challenges and Limitations

Kruti is a first-version product in an enormously ambitious category. For now, its capabilities are limited to a few partners, which may slow its adoption. Real-world agentic AI, systems that autonomously execute multi-step tasks across third-party services, is one of the hardest engineering problems in the field. Reliability, error handling, and trust-building with users will be determinant factors in whether Kruti achieves widespread adoption or remains a compelling demo.

These challenges are not unique to Kruti. Any business deploying conversational AI at scale will encounter the same fundamental tension between automation ambition and real-world reliability. A clear-eyed understanding of the genuine risks and operational downsides that come with chatbot deployment is essential context for evaluating any agentic AI system — including this one.

The monetization question is also open. Future monetization is expected to come through transaction commissions or premium subscriptions. Giving away image generation and research tools for free builds a user base; converting that base into sustainable revenue while keeping the product accessible is a balance the company has not yet publicly resolved.

Measuring whether the product is actually delivering value, for users and for the business — will require robust analytics frameworks that track how chatbot interactions translate into real outcomes. This is an area where many first-generation AI deployments fall short, and Kruti will be no exception to that challenge.

Who Should Pay Attention to Kruti?

Kruti AI is not a ChatGPT competitor in the conventional sense, it is a different category of product pursuing a different user need in a different market. Its significance is less about raw model capability today and more about the infrastructure, intent, and trajectory it represents: India building AI that is Indian by design, not adapted from Western foundations as an afterthought.

For Indian users, developers, and businesses, Kruti is worth watching closely. The agentic AI model it is pursuing, where a single interface replaces a constellation of apps, is where the global industry is heading. The broader shifts in how AI is reshaping sales, support, and customer engagement in 2025 suggest that platforms built around task execution rather than conversation are becoming the new competitive baseline.

Krutrim may be building that future for India before the global players get there first.