Deepgram

Converts live audio streams to text via WebSocket protocol using Flux English or Flux Multilingual models optimized for conversational speech. Implements automatic turn-taking detection to identify speaker transitions in real-time, enabling natural voice agent interactions without explicit end-of-speech markers. Processes continuous audio streams with sub-100ms latency targets for conversational responsiveness.

Processes pre-recorded audio files via REST API using Nova-3 Monolingual or Nova-3 Multilingual models to generate full transcripts with speaker identification, automatic punctuation, capitalization, and readability enhancements. Supports multi-channel audio for automatic speaker attribution. Returns structured JSON with word-level timing, confidence scores, and speaker labels for each utterance.

Deepgram offers both cloud-hosted API and self-hosted deployment options, allowing organizations to run speech-to-text and text-to-speech models on their own infrastructure. Self-hosted deployments provide data residency guarantees and eliminate data transmission to Deepgram's servers, addressing privacy and compliance requirements.

Deepgram offers a free tier providing $200 in usage credits with no expiration date, allowing developers to experiment with all API features without payment. Free tier includes concurrency limits (50 STT REST, 150 STT WebSocket, 45 TTS, 10 Audio Intelligence) but no per-minute or per-hour request rate limits. No credit card required for signup.

Deepgram offers two primary pricing models: pay-as-you-go with per-minute rates for STT and TTS, and Growth plan with annual pre-paid credits offering up to 20% discount. Pricing varies by model (Flux vs. Nova-3) and processing mode (streaming vs. batch). Enterprise plans available with custom pricing and concurrency limits.

Interactive web interface allowing developers to test Deepgram APIs without writing code. Supports uploading audio files, configuring model parameters, and viewing real-time transcription results with detailed metadata (confidence scores, timing, speaker attribution). Provides visual feedback and API request/response inspection for learning and debugging.

Rate limiting enforced via concurrent connection limits rather than requests-per-second, with different quotas for each API endpoint and pricing tier. STT streaming supports 150 concurrent WSS connections (Free), 225 (Growth); REST API supports 100 concurrent; TTS supports 45-60 concurrent; Audio Intelligence supports 10 concurrent. Enables predictable scaling for applications with variable request patterns.

Four-tier pricing model: Free tier with $200 credit (no expiration), Pay-As-You-Go with per-minute pricing ($0.0058-$0.0165/min for STT depending on model), Growth tier with annual commitment ($4,000+ minimum, up to 20% discount), and Enterprise tier with custom pricing. Enables organizations to start free and scale to enterprise volumes with predictable costs.

Automatically identifies the language spoken in audio and transcribes it using Nova-3 Multilingual model supporting 45+ languages, or uses Flux Multilingual for real-time streaming across 10 languages. For streaming conversations, Flux Multilingual can handle language switching within a single session without requiring manual language specification or model switching.

Biases transcription toward domain-specific terminology by accepting a list of keywords or phrases that should be prioritized during decoding. The model adjusts its language model weights to favor these terms, improving accuracy for technical jargon, proper nouns, product names, or industry-specific vocabulary that might otherwise be misrecognized.

Deepgram offers custom model training for organizations with proprietary audio data, domain-specific vocabulary, or unique acoustic environments. Custom models are trained on client-provided datasets to optimize accuracy for specific use cases, languages, or speaker populations. Pricing and training timeline available through enterprise sales.

Converts text input to natural-sounding speech using Deepgram's Speak model, supporting multiple voices and languages. Implements streaming output via WebSocket or HTTP chunked transfer, enabling real-time audio playback without waiting for full synthesis completion. Supports continuous text stream processing for applications that generate text incrementally (e.g., LLM outputs).

Voice Agent API provides a single endpoint that orchestrates speech-to-text transcription, routes to external LLMs or internal logic, and synthesizes responses back to speech. Handles conversation state management, turn-taking, interruption detection, and automatic language detection within a single WebSocket connection. Abstracts away the complexity of coordinating multiple models and managing real-time audio streams.

Audio Intelligence API analyzes transcribed speech to extract emotional tone (sentiment analysis) and identify subject matter (topic detection). These analyses are performed on transcripts after speech-to-text processing, providing structured metadata about conversation content and speaker emotion. Supports batch processing of multiple transcripts.

Command-line interface providing direct access to all Deepgram API endpoints without writing code. Includes 28 pre-built commands for STT, TTS, and Audio Intelligence operations. Implements a Model Context Protocol (MCP) server, enabling AI agents and LLMs to invoke Deepgram capabilities as structured tools with schema-based function calling.

Deepgram provides native SDKs for five major programming languages, each implementing the full API surface (STT, TTS, Audio Intelligence, Voice Agent). SDKs handle authentication, request/response serialization, WebSocket connection management, and error handling. Abstracts API details while maintaining language-specific idioms and conventions.

Deepgram provides an enterprise-grade API for speech-to-text and text-to-speech, leveraging advanced deep learning models for high accuracy and real-time processing, ideal for applications requiring transcription and audio generation.