Streaming and prerecorded transcription
Nova speech recognition models transcribe live streams and recorded audio. Available capabilities include speaker diarization, automatic language detection, and formatting for punctuation, dates, and currency.
Deepgram
Speech-to-text, text-to-speech, and voice agent APIs for developers and enterprises.
O que é Deepgram?
Deepgram is a voice AI platform for developers, product teams, and enterprises building applications around spoken conversations. It provides APIs for speech-to-text, text-to-speech, voice agents, and audio intelligence, with support for live audio and prerecorded transcription.
Teams can integrate individual speech services or use the unified Voice Agent API to combine recognition, speech generation, and language model orchestration. Enterprise options include custom speech models and deployment choices for different infrastructure and data requirements.
O que Deepgram pode fazer?
Conversation-aware speech recognition
Flux STT includes turn detection and interruption handling for real-time voice agents. Flux Multilingual supports multiple languages within a single conversation.
Conversational speech synthesis
Flux TTS generates speech for live conversations, maintaining tone, expression, and continuity across the conversation.
Unified voice agent orchestration
The Voice Agent API combines STT, TTS, and LLM orchestration. It maintains conversation history and supports function calling, prompt updates, and response injections to connect agents with application logic and external systems.
Transcript enrichment and redaction
Speech-to-text add-ons can identify entities such as names, organizations, and locations, redact sensitive personal information, and use keyterm prompting to improve recognition of domain terminology.
Custom speech models
Enterprises can commission speech-to-text models trained on proprietary or novel datasets for specialized vocabulary and edge-case audio scenarios.
Casos de uso
“Customer service voice agents”
Build agents that converse with callers, retrieve information from back-office systems, and execute domain-specific actions through application logic.
“Meeting transcription products”
Use transcripts as the input for meeting-note applications, with speaker labels and readable formatting helping organize conversations.
“Recorded conversation analysis”
Process archived calls or other recordings to extract named entities and prepare redacted transcripts for downstream analysis.
“Voice capabilities for platforms”
Embed speech recognition and speech generation into communications or conversational AI platforms that serve other developers and businesses.
Perguntas frequentes
How does pricing work?
API pricing depends on the selected speech service, model, usage, and add-ons. Free evaluation access is advertised; check current credit allowances, any expiry, prepaid commitments, discounts, and enterprise terms before use.
What deployment options are available?
Deepgram supports cloud and self-hosted deployments. Discuss dedicated runtime options and regional endpoint availability with the provider for your infrastructure and data requirements.
Which languages does transcription support?
The pricing page lists 45+ languages for Nova models. Language and feature availability vary by model; Flux offers English and multilingual options for conversational recognition.
Does the Voice Agent API provide the audio transport?
The architecture uses a partner-provided transport layer to stream audio into Deepgram and play generated speech back to users. Telephony partners can connect through PSTN or SIP; your application supplies business logic and integrations.
Informações rápidas
- Product category
- Voice AI API platform
- Primary users
- Developers, product teams, platforms, and enterprises
- Interface
- REST and WebSocket APIs
- Website
- https://deepgram.com
Análise de tráfego de Deepgram
Os dados de tráfego são apenas para referência.
- Classificação do domínio
- 80
Alternativas ao Deepgram
Speechmatics
www.speechmatics.com
Speechmatics provides speech-to-text APIs for pre-recorded and real-time audio, with support for multilingual speech, multiple speakers, and voice-agent workflows. It is designed for teams building transcription and Voice AI applications.
Lemonfox.ai
lemonfox.ai
Speech-to-text and text-to-speech APIs for developers building audio workflows.
SpeechifyAI
speechify.ai
SpeechifyAI is a developer API for expressive text-to-speech and voice cloning. It provides streaming Simba models, catalog and cloned voices, SSML support, and a free starting tier for building speech into applications.
4ALL API
4allapi.com
4ALL API is an API aggregation gateway for enterprises and developers that provides one access layer for text, image, video, and audio models from multiple providers. It supports OpenAI-compatible access, usage-based billing, model routing, and failover workflows.
Inworld AI
inworld.ai
Realtime voice, speech, and language-model infrastructure for consumer applications.
Typecast
typecast.ai
AI voice generation and video editing for creators, developers, and teams.