
by Gladia
Gladia is an audio infrastructure API that transcribes speech in real time or from uploaded files and adds audio intelligence such as sentiment and entity detection. It is aimed at developers building voice products.
APIUsage-based
Step 1: Create an account and use the starting credits.
Step 2: Try audio in the Playground or read the documentation.
Step 3: Send audio by REST upload for async, or WebSocket for real-time.
Step 4: Enable features such as diarization, summarization or PII redaction in your request.
Step 5: Receive the transcript and enrichment data, and send it onward with webhooks or integrations.
Step 6: Top up credits or contact sales for Enterprise.
On Self-serve, training is opted in by default and you can opt out at no cost from your profile. Enterprise is opted out by default.
Self-serve allows 30 concurrent real-time sessions and 25 concurrent pre-recorded jobs. Enterprise removes the limits, and requests above the limit get a 429 response.
Real-time returns results as audio streams in over a live connection. Async processes uploaded files and is the cheaper route when a short delay is acceptable.
Gladia lists 100% EU data residency on its home page and also compares EU and US hosting options.
Resemble AIResemble AI detects AI-generated audio, video and images and offers watermarking and identity tools. It is used for fraud prevention, meeting security and content moderation, through an API or built-in integrations.
SpeechmaticsSpeechmatics provides APIs for speech to text, text to speech and voice agents. It offers real-time and batch transcription in the cloud or on your own infrastructure.
OpenAI Realtime APIOpenAI's Realtime API lets developers build speech-to-speech voice agents that work directly with audio, keep conversation state and call tools. It is for developers building voice assistants and other low-latency conversational apps.