Gladia

Gladia is an audio infrastructure platform that transcribes and enriches speech through a single API. It is built for developers making voice products such as meeting assistants, contact center software, voice agents and media tools. The speech to text API works in real time over streaming connections and asynchronously on uploaded files or video, with speaker diarization, automatic language detection and switching, and support for more than 100 languages. Beyond transcripts, an audio intelligence layer extracts structured data from conversations. The vendor lists its Solaria 3 model at 9.6 percent word error rate on real English audio and states 100 percent EU data residency, a 99.95 percent uptime SLA, and GDPR, HIPAA and SOC 2 Type 2 compliance. A hosted playground, documentation, a changelog and a blind test tool for comparing speech to text models are available, along with GladiaFlow, an app built on the platform. Pricing is per hour of audio. The Starter plan is pay as you go at $0.61 per hour for async and $0.75 per hour for real time transcription, with 50 euros in free credits, 30 concurrent real time requests and 25 concurrent async requests. The Growth plan lowers rates to as little as $0.20 per hour async and $0.25 per hour real time for teams that commit upfront, and is arranged through the sales team.