LALAL.AI
Audio & musicSeparate vocals and instruments from existing recordings, with additional voice-cleaning and audio tools.
Deepgram offers APIs for real-time and recorded speech-to-text, text-to-speech and voice agents. New accounts get $200 of free credit with no card.
Deepgram is voice AI infrastructure for developers. It provides the pieces you need to build products that listen and talk: speech-to-text for live streams and recorded files, text-to-speech voices, and a Voice Agent API that joins listening, a language model and speech into one agent. Its own description is “the real-time infrastructure behind speech-to-text, text-to-speech, and voice agents that listen and respond the way people actually talk.” Pricing is usage-based, billed per minute of audio or per thousand characters, with no minimum.
Editorial suggestion: compare it on your own audio before choosing. Accents, product names and background noise change results more than any benchmark, and the price difference between streaming and pre-recorded transcription matters at volume. The $200 starter credit is enough for a meaningful test of several models.
Call centres
Voice agents
Meeting tools
Media
Official product image. Click to inspect the details.


Official examples and public-page captures; these are not results from a paid-feature trial.
Nova-3 models for streaming and pre-recorded audio in monolingual and multilingual versions, Flux for real-time conversation, and Whisper Large for recorded files.
Aura-2, Aura-1 and Flux TTS voices; Deepgram lists more than 45 languages and over 100 voice profiles.
One API that listens, thinks and speaks, with function calling, for example to book appointments; you can bring your own LLM or TTS.
Enterprise plans cover data and deployment requirements, volume pricing and support; every tier runs on the same backend.
Sign up
Create an API key
Test recorded audio
Move to streaming
Create an account on deepgram.com to receive the $200 credit; no card is required.
Generate a key in the console and keep it on your server, not in front-end code.
Send a few of your own recordings to the pre-recorded endpoint and compare models.
Connect a live microphone or call stream and measure latency before choosing a plan.
Pay as you go has no minimum and no expiry. Examples: Nova-3 Monolingual costs $0.0077 per minute streaming or $0.0043 per minute pre-recorded; Flux STT English $0.0077 per minute; Aura-2 $0.030 per 1,000 characters; the Voice Agent API $0.075 per minute in its standard configuration.
The Growth plan starts at $4,000 a year in prepaid credit and lowers rates by up to 20%. Enterprise pricing is custom.
New accounts get $200 of credit with no card; after that you pay per minute or per thousand characters.
Nova-3 for most transcription, Flux for real-time conversations such as voice agents.
Nova-3 has multilingual versions, and Deepgram lists 45+ languages for text-to-speech.
Yes. The Voice Agent API has configurations to bring your own LLM, your own TTS, or both.
Usually yes: for Nova-3 Monolingual, streaming is $0.0077 and pre-recorded $0.0043 per minute on pay as you go.
The Growth plan, from $4,000 a year prepaid, cuts rates by up to 20%; Enterprise is custom.
Explore platforms, inputs and outputs, licensing, and access requirements.
These tools overlap with parts of the workflow; compare the specific task, export requirements and current plan rather than treating them as identical products.
Audio you send is processed by Deepgram, so check its data processing terms and, for regulated data, the enterprise options before sending real customer calls. Keep API keys on the server and limit which team members can create them.
Loading experiences…
Bring a real task and see how it fits the way you work.
User experiences