Deepgram API

Name: Deepgram API
Brand: Deepgram API
Availability: InStock

✓ Official Vendor SpecAI/MLSpeechapiKey, bearer39 EndpointsREST

For Agents

Transcribe audio to text, synthesize speech from text, and analyze language content. Supports 30+ languages with speaker diarization, topic detection, and custom vocabulary for domain-specific accuracy.

Install Jentic One Beta

Connect the Deepgram API to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Deepgram API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with Deepgram API.

Transcribe pre-recorded audio files with speaker diarization, punctuation, and paragraph segmentation

Convert text to natural-sounding speech with multiple voice models and output formats

Analyze text for intent, sentiment, topic detection, and summarization via the Read endpoint

GET STARTED

Start building with Deepgram API

Explore with Jentic

View OpenAPI Document

Use Cases

Patterns agents use Deepgram API for, with concrete tasks.

★ AI Agent Audio Transcription Pipeline

AI agents transcribe audio content (meetings, calls, podcasts, voicemails) into structured text through Jentic. The agent posts audio to the Listen endpoint with parameters for diarization, punctuation, and paragraphs, and receives timestamped text with speaker labels. Jentic handles API key injection and model selection, so the agent focuses on processing the transcript output for downstream tasks like summarization or search indexing.

POST an audio file URL to /v1/listen with parameters diarize=true, punctuate=true, and model=general, then extract the transcript text with speaker labels from the response

Text-to-Speech Generation

Generate natural-sounding speech audio from text input for voice assistants, accessibility features, or content narration. The Speak endpoint accepts plain text or SSML and returns audio in multiple formats (WAV, MP3, OGG). Multiple voice models provide different speaking styles and tones. Response streaming enables real-time audio playback as generation progresses.

POST text content to /v1/speak with a specified voice model and receive the generated audio file in MP3 format

Meeting and Call Analytics

Extract actionable intelligence from meeting recordings and phone calls with speaker-attributed transcription, topic detection, and summarization. Deepgram's meeting-optimized model handles overlapping speech, filler word removal, and domain terminology. The output includes word-level timestamps enabling precise navigation to specific moments in recordings and automated meeting minutes generation.

Transcribe a meeting recording via POST /v1/listen with model=meeting, diarize=true, summarize=true, and topics=true, then parse the response for speaker turns and topic segments

Document and Text Understanding

Analyze written text for intent, sentiment, topics, and summaries through the Read endpoint. This enables processing of transcripts, emails, support tickets, and documents without audio input. The API returns structured analysis including detected topics with confidence scores, overall sentiment, and concise summaries suitable for search indexing or automated routing.

POST a text document to /v1/read with intents=true and summarize=true, then extract the detected intents and summary from the response

Key Endpoints

39 endpoints — transcribe audio to text, synthesize speech from text, and understand language through 39 endpoints covering deepgram's ai speech platform.

METHOD

PATH

DESCRIPTION

POST

/v1/listen

Transcribe audio to text with model and language options

POST

/v1/speak

Synthesize text to speech audio

POST

/v1/read

Analyze text for intent, sentiment, and topics

GET

/v1/models

List available transcription models

GET

/v1/projects

List all projects for the account

GET

/v1/projects/{project_id}/usage

Get usage statistics for a project

POST

/v1/projects/{project_id}/keys

Create a new API key for a project

POST

/v1/auth/grant

Generate a temporary JWT for scoped access

POST

/v1/listen

Transcribe audio to text with model and language options

POST

/v1/speak

Synthesize text to speech audio

POST

/v1/read

Analyze text for intent, sentiment, and topics

GET

/v1/models

List available transcription models

GET

/v1/projects

List all projects for the account

Why Jentic?

Three things that make agents converge on Jentic-routed access.

Credential isolation

Your Deepgram API key is stored once, encrypted, by your own Jentic One instance and applied as the Authorization: Token header at execution time. It never enters the agent's prompt, logs, or context.

Intent-based discovery

Agents search Jentic by intent such as 'transcribe an audio recording' or 'synthesize speech from text', and Jentic returns the matching Deepgram Listen, Speak, or Read operation with its parameter schema so the agent selects the right model and options without navigating documentation.

Time to first call

FAQs

Specific to using Deepgram API through Jentic.

What authentication does the Deepgram API use?

The Deepgram API uses API key authentication passed in the Authorization header with a 'Token' prefix (Authorization: Token <API_KEY>). It also supports JWT bearer tokens for temporary scoped access. API keys are scoped to projects and can have different permission levels. Through Jentic, API keys are stored encrypted in your Jentic One instance - agents receive pre-authenticated headers without the raw key entering their context.

Can I transcribe audio with speaker identification?

Yes. Add diarize=true to your POST /v1/listen request and Deepgram returns speaker labels for each word and utterance in the transcript. The diarization model distinguishes between speakers in multi-person recordings like meetings and interviews. Each word in the response includes a speaker field with a numeric speaker identifier.

What are the rate limits for the Deepgram API?

Deepgram's rate limits depend on your plan tier and concurrency level. The API supports concurrent transcription requests with limits based on your subscription. Pay-as-you-go plans start with a set concurrency level that scales with usage. The API returns HTTP 429 with rate limit headers when the concurrency ceiling is reached.

How do I transcribe an audio file through Jentic with an AI agent?

Install the Jentic SDK with pip install jentic, then search for 'transcribe audio to text'. Jentic returns the POST /v1/listen operation schema with supported parameters (model, language, diarize, punctuate, paragraphs, summarize). The agent posts the audio URL or binary, and Jentic injects the Authorization: Token header - the API key never appears in the agent's context.

Which transcription models does Deepgram offer?

Deepgram offers domain-optimized models including 'general' for broad use, 'meeting' for multi-speaker meetings, 'phonecall' for telephony audio, and 'voicemail' for short-form messages. Each model is tuned for its domain's acoustic characteristics and vocabulary. You can list available models via GET /v1/models and select per-project models via GET /v1/projects/{project_id}/models.

Does Deepgram support text-to-speech generation?

Yes. The POST /v1/speak endpoint accepts text input and returns synthesized audio in formats like WAV, MP3, and OGG. You select a voice model in the request to control the speaking style and tone. The endpoint supports streaming responses for real-time playback applications.

Can I limit what my agent is allowed to do with the Deepgram API?

Yes. Because Jentic One is self-hosted, you decide which Deepgram operations your agent may call, so you can grant read-only transcription and synthesis through POST /v1/listen, POST /v1/speak, and POST /v1/read while withholding state-changing operations like POST /v1/projects/{project_id}/keys or POST /v1/auth/grant. Deepgram puts the project id in the URL path, as in /v1/projects/{project_id}/usage, so your rules can pin the agent to a single project. Your Deepgram API key is stored encrypted by your own instance and applied at execution time, so it never enters the agent's prompt or context.

Deepgram API

Connect the Deepgram API to your agent

Capabilities

Start building with Deepgram API

Use Cases

Key Endpoints

Why Jentic?

Related APIs

FAQs