KeenASR SDK 2.0: Smarter, Faster, and Now on the Web

Author: Ognjen Todic | September 23, 2025

At Keen Research, we’ve always believed that powerful speech recognition should be fast, private, and accessible across devices. In late December, we launched KeenASR SDK 2.0, a major upgrade that expanded capabilities on iOS, Android, and Linux. Now, with our latest update – v2.0.4 – we’re taking another big step: KeenASR SDK is available for the Web.

This means the same on-device accuracy and privacy that innovators in EdTech, Enterprise, Healthcare, and other industries already rely on can now run directly in the browser — opening up entirely new possibilities for voice-enabled experiences.

What’s New in KeenASR SDK 2.0

Version 2.0 introduced a host of enhancements designed to make speech recognition more powerful and flexible:

Oral Reading tools: Contextual decoding graphs, alternative pronunciations, and robust handling of out-of-vocabulary words allow developers to build richer and more accurate reading experiences.
Cross-platform Dashboard support: Developers can leverage Dashboard, during development, on both iOS and Android platforms
Goodness of Pronunciation scoring: Phoneme-level scores in the English bundle enable real-time learner feedback, formal assessments, and personalized learning paths.
Multi-language support: In addition to English (adult & children models) and Spanish, we now support German and French.

And now, with 2.0.4, there’s one more milestone:

Web SDK: KeenASR runs directly in the browser — no servers, no plugins, no backend. Everything is processed locally via a lightweight JavaScript library.

Web SDK: Voice in the Browser

For the first time, developers can integrate KeenASR into browser-based apps as easily as they would on mobile platforms. This opens the door for more interactive and accessible voice-enabled experiences — whether in education, retail, or frontline workflows.

Web SDK also makes demos and prototypes easier to share. We’ve already built various sample applications including oral reading assessment, pronunciation scoring, and warehouse voice-picking; you can try KeenASR demos here.

Real-World Impact Across Verticals

From classrooms to warehouses to immersive virtual worlds, KeenASR SDK 2.0 is already shaping how organizations bring voice into their products and workflows:

EdTech: From oral reading assessment to interactive story apps, 2.0 enables fun, effective, and scalable learning solutions.
Enterprise & Frontline: Faster, offline-ready voice interfaces for order picking, inspections, maintenance, and safety checklists.
Retail: Private, real-time voice interactions in mobile shopping apps, kiosks, or virtual assistant for retail employees.
XR: Immersive AR/VR training powered by fast, on-device recognition.

Get Started with 2.0

Whether you’re developing for mobile, custom hardware, or the web, KeenASR SDK 2.0 gives you more flexibility, accuracy, and performance than ever before.

Try the SDK now

Read the full release notes

Stay Connected

Want to stay up to date with the latest product updates, customer stories, and insights on on-device speech recognition?