Speechmatics is a Voice AI company that provides highly accurate, real-time speech-to-text and text-to-speech APIs supporting over 55 languages. Their technology is built on deep neural networks and is designed to understand diverse accents and dialects, making it one of the most inclusive speech recognition platforms. The company offers a speech intelligence suite including summarization, sentiment analysis, and topic detection, and is compliant with GDPR, HIPAA, and SOC 2. With $7.5M in funding, Speechmatics serves enterprises across media, healthcare, contact centers, and education.
Speechmatics has opened early access for its on-device speech-to-text engine, which runs locally on macOS and Windows to provide privacy and predictable costs. The technology supports 55+ languages and is designed for enterprise ISVs and OEMs requiring offline or secure transcription.
Speechmatics has updated its Melia 1 multilingual speech-to-text model to improve code-switching accuracy for Arabic, Mandarin, and Tamil. The company claims the model achieves a 15.1% mixed error rate on Arabic-English code-switching and provides faster batch processing speeds.
Speechmatics has integrated its Linden speech-to-text model into the LiveKit Inference platform, allowing developers to access the model without separate API keys or billing. The integration aims to improve transcription accuracy for voice agents in noisy, multi-speaker, and non-native language environments.
Speechmatics published a detailed post on how it rebuilt Adobe Premiere’s on-device speech engine, saying its model outperformed Whisper on the same hardware and was optimized to run on consumer laptops.
Speechmatics details the engineering process behind updating its speech-to-text engine for Adobe Premiere to maintain competitive performance against OpenAI's Whisper model on consumer hardware. The article focuses on technical techniques like quantization, custom model optimization scripts, and GPU acceleration to manage resource constraints on personal computers.
Speechmatics has launched Melia, a new multilingual speech-to-text model capable of handling code-switching across 56+ languages. The model is available in production preview via batch processing and is positioned as a cost-effective solution for contact center analytics and broadcast captioning.
NCI has implemented Speechmatics' advanced speech recognition technology to enhance its real-time captioning services. The partnership aimed to revolutionize captioning by addressing existing challenges and improving efficiency. The article highlights the solutions implemented and the positive outcomes of this collaboration between the two companies.
Media Track, a global media monitoring company, has enhanced its broadcast and print transcription accuracy and scalability by integrating Speechmatics' real-time speech-to-text technology. This partnership allows Media Track to improve its global media monitoring capabilities.
AI-Media is enhancing its live captioning capabilities by integrating Speechmatics' advanced transcription engine. This collaboration aims to deliver accurate, scalable, and multilingual real-time captions for global media and educational content.
Speechmatics and Adobe have collaborated to introduce cloud-grade speech recognition directly on-device within Adobe Premiere. This new feature enables private, offline, and accurate speech-to-text capabilities for professional creators using Premiere.
Will Williams
Chief Technology Officer
