News Hub

Coval introduces TTS latency benchmark called Perceived TTFA.

Gradium TTS is available on LiveKit Inference.

Gradium enables creating new voices in seconds through prompt-based voice design.

Gradium TTS balances latency and accuracy tradeoffs in text-to-speech synthesis.

Gradium voices are now available on AudioStack platform.

New voices: how we pick the one that wins.

Gradium enables speech-to-text systems to recognize custom vocabulary through keyword boosting.

Gradium offers natural conversation search with live retrieval capabilities.

Gradium Translate achieves better accuracy-latency tradeoff than Gemini 3.5 Live Translate and GPT Realtime Translate.

Gradium released an upgraded text-to-speech system with improved accuracy.

Gradium introduced semantic voice activity detection using meaning rather than silence.

Phonon achieved 1.00% WER on Seed-TTS, smaller than competing models.

Gradium achieved top ranking on Coval text-to-speech benchmarks.

Gradium Voice launches on AWS as SaaS and SageMaker model image.

Gradium claims most accurate multilingual text-to-speech by numerical metrics.

Gradbot enables voice agents in 50 lines of code.

Phonon is Gradium's text-to-speech model optimized for edge devices.

Gradium measures and reduces latency in text-to-speech systems for voice agents.

Gradium optimizes quality versus latency tradeoffs in real-time text-to-speech models.

Gradium launched a startup program for building voice agents.