Shunya Labs Zero Codeswitch speech recognition model launched

India’s linguistic diversity has long posed a challenge for voice-based technologies, particularly in regions where everyday conversations blend multiple languages within a single sentence. Addressing this gap, Shunya Labs has announced the launch of the Shunya Labs Zero Codeswitch speech recognition model, a foundation model designed to natively understand code-mixed and multilingual Indian speech.

The model builds on Shunya Labs’ recent performance on the OpenASR leaderboard, where it achieved a 3.10% Word Error Rate, representing a 48% improvement over the next-best competitor. Zero Codeswitch is engineered to run efficiently on standard CPUs, reducing deployment costs by up to 20 times while maintaining sub-100 millisecond latency for real-time use cases.

Designed for India’s code-mixed speech patterns

Most existing speech recognition systems are optimised for single-language inputs. In contrast, everyday communication in India frequently combines Hindi, English, and regional languages, often using transliteration, informal grammar, and colloquial expressions.

Zero Codeswitch is designed to process such speech natively, without relying on intermediate translation layers that convert Indian languages into English. Instead, it uses a unified architecture capable of generating mixed Hindi and English tokens, enabling accurate transcription of natural phrases commonly used in daily conversations.

Foundation models trained on Indian speech data

Unlike global speech models that are primarily trained on English datasets and later adapted for Indian languages, Shunya Labs’ foundation models are trained from the ground up on Indian speech data.

Zero Codeswitch has been trained on millions of hours of real-world Indian audio, capturing variations in accent, dialect, pronunciation, slang, and transliteration. While optimised for Hinglish and other code-mixed speech patterns, the model is part of a broader Shunya Labs platform that supports more than 200 languages, including over 40 Indian languages.

Use cases across regulated and public sectors

The company expects the Shunya Labs Zero Codeswitch speech recognition model to serve as core infrastructure for voice-based applications across sectors such as fintech, healthcare, and government services. These use cases are particularly relevant in regions where English proficiency is limited and voice remains the most accessible digital interface.

Designed for enterprise and public-sector environments, Zero Codeswitch supports on-premises and air-gapped deployments. This allows organisations to train domain-specific variants while retaining full control over sensitive data. The model supports CPU-first deployments, reducing reliance on specialised GPU infrastructure.

Shunya Labs maintains compliance with HIPAA, SOC 2 Type II, and ISO 27001 standards, addressing data privacy and security requirements for regulated industries.

Leadership perspective

Ritu Mehrotra, CEO and Co-Founder, Shunya Labs, said the company has focused on deep research rather than short-term narratives, with Zero Codeswitch aimed at building foundational AI technology for Indian languages that prioritises accuracy, latency, and real-world usability.

Sourav Bandyopadhyay, CTO and Co-Founder, Shunya Labs, said the model reflects the company’s first-principles approach, describing Zero Codeswitch as an intelligence layer engineered for India’s linguistic diversity and real-world deployment conditions.

Growing domestic AI infrastructure

Headquartered in Gurugram, Shunya Labs is part of a growing ecosystem developing domestic intellectual property in core AI technologies. The company’s models are benchmarked against global competitors and are already being adopted by customers in India and internationally.

Availability

The Shunya Labs Zero Codeswitch speech recognition model is available for enterprise pilots and production deployment. Organisations can request access through the company’s website.

Read More:

FAIITA's Navin Gupta at FITAG Tech Expo: IT channel ecosystem must evolve

Partner Pulse: Digi Edze Soft Solutions | Managed Service Provider (India)

Technology predictions 2026: how agentic AI will reshape enterprises

FITAG Tech EXPO Day 2: Cybersecurity threats took centre stage