Cross-platform offline speech recognition plugin for Unreal Engine. Convert speech to text with advanced Whisper AI technology across all platforms without internet connection.
Runtime Speech Recognizer provides a comprehensive system for real-time speech recognition, featuring offline processing powered by Whisper OpenAI technology. From recognizing commands to transcribing full conversations, handle all your speech recognition needs with a single solution that works entirely offline.
Works entirely on-device without requiring internet connection
Choose from Tiny, Base, Small, Medium, or Large models to balance accuracy and performance
Recognize speech in over 95 languages with automatic language detection
Vulkan-based acceleration on Windows for significantly faster recognition
Process audio in real-time as it's being captured, ideal for interactive applications and voice commands.
Process complete audio files or buffers in a single operation for maximum accuracy.
Combine with Voice Activity Detection for optimal recognition of speech segments and command-based interfaces.
Powered by Whisper OpenAI technology, specifically the optimized whisper.cpp implementation, providing state-of-the-art speech recognition with:
Floating point 32-bit interleaved PCM audio format
Works with any audio source that can provide PCM data, including Runtime Audio Importer
Process audio as it's being captured or in complete chunks
Voice Activity Detection recommended for streaming scenarios
Runtime Speech Recognizer works seamlessly with other plugins to create complete voice interaction solutions for your Unreal Engine projects.
Capture microphone input with Voice Activity Detection to provide clean audio segments for optimal speech recognition performance.
Learn moreCreate realistic lip sync for MetaHuman characters that speak the responses to recognized speech, enabling natural conversational interfaces.
Learn moreComplete the voice interaction loop by generating spoken responses to recognized speech commands with offline TTS technology.
Learn moreProcess recognized speech with AI models from OpenAI, Claude, or DeepSeek to create intelligent conversational agents in your applications.
Learn moreCombine Runtime Audio Importer for voice capture, Runtime Speech Recognizer for speech-to-text, Runtime AI Chatbot Integrator for intelligent responses, and Runtime Text To Speech for voice output to create a fully functional voice assistant or NPC conversation system.
Get started quickly with our detailed documentation and receive support through multiple channels. From basic usage to advanced speech recognition techniques, we're here to help you succeed.
Step-by-step guides for all features
Visual guide to get you started quickly
Get real-time help from developers and users
Direct contact for custom development needs
Get started quickly with the included demo project that showcases all the key features of Runtime Speech Recognizer.
Runtime Speech Recognizer supports over 95 languages, making it suitable for global applications.
Turkish, Polish, Dutch, Swedish, Finnish, Vietnamese, Ukrainian, Greek, Czech, Romanian, Danish, Hungarian, Norwegian, Thai, Croatian, Bulgarian, Lithuanian, Welsh, Slovak, Persian, Latvian, Bengali, Serbian, Slovenian, Estonian, and many more.
View complete list of 95+ supported languagesEnable natural voice interaction in your applications with offline, cross-platform speech recognition. From voice commands to full conversation transcription, unlock new possibilities for user interaction.
