# Deepgram

Deepgram provides AI-powered speech recognition and understanding services, offering APIs for real-time and pre-recorded audio transcription, text-to-speech, and audio intelligence.

- **Category:** artificial intelligence
- **Auth:** API_KEY
- **Composio Managed App Available?** N/A
- **Tools:** 11
- **Triggers:** 0
- **Slug:** `DEEPGRAM`
- **Version:** 20260316_00

## Tools

### Get Model by ID

**Slug:** `DEEPGRAM_GET_MODEL`

Retrieve metadata for a specific Deepgram model by its UUID. Returns detailed model information including name, architecture, supported languages, version, and capabilities. Works for both STT (speech-to-text) and TTS (text-to-speech) models.

### Get Public Models

**Slug:** `DEEPGRAM_GET_MODELS`

Retrieve metadata on all public Deepgram models (speech-to-text and text-to-speech). Returns comprehensive model information including supported languages, architectures, versions, and capabilities. Set include_outdated to True to include deprecated versions.

### List Deepgram Projects

**Slug:** `DEEPGRAM_GET_PROJECTS`

Tool to list all Deepgram projects. Use after authenticating with your API key.

### Get Project Usage Summary

**Slug:** `DEEPGRAM_GET_PROJECT_USAGE_SUMMARY`

Retrieves aggregated usage statistics for a Deepgram project including total audio duration, billable duration, number of requests, channels processed, and confidence/relevance scores. Returns both overall totals and breakdowns by model/accessor/tag. Use this to analyze API consumption, track costs, or monitor transcription quality metrics over time.

### Get Public TTS Models

**Slug:** `DEEPGRAM_GET_PUBLIC_TTS_MODELS`

Tool to fetch metadata about all latest public TTS voice models. Use when you need to list available Deepgram TTS voices.

### List Project Scopes

**Slug:** `DEEPGRAM_LIST_PROJECT_SCOPES`

Tool to list all scopes for a specified Deepgram project. Use when you need to retrieve all permission scopes for a project.

### List Think Models

**Slug:** `DEEPGRAM_LIST_THINK_MODELS`

Tool to list available think models for AI agent processing and voice agent configuration. Use when you need to see which think models are available for voice agents.

### Transcribe Pre-recorded Audio

**Slug:** `DEEPGRAM_SPEECH_TO_TEXT_PRE_RECORDED`

Tool to transcribe pre-recorded audio files into text. Use when converting a publicly accessible audio file URL to text. Primary transcript is at `results.channels[0].alternatives[0].transcript` in the response. Silent audio returns a valid empty transcript, not an error. Verify supported models and language codes via `DEEPGRAM_GET_MODELS` when uncertain.

### Summarize Audio

**Slug:** `DEEPGRAM_SUMMARIZE_AUDIO`

Generate a concise text summary from audio content using Deepgram's AI-powered summarization. This tool transcribes audio from a publicly accessible URL and generates a summary of the content. The summary captures key points and main ideas from the audio, making it ideal for: - Meeting recordings - Podcast episodes - Interview recordings - Voice messages - Any spoken content that needs quick understanding Requirements: - Audio file must be publicly accessible via HTTP/HTTPS URL - Supported formats: WAV, MP3, MPEG, FLAC, WEBM - Returns both full transcript with word-level timing and a generated summary Note: The audio will be downloaded and sent to Deepgram for processing.

### Text-to-Speech (REST)

**Slug:** `DEEPGRAM_TEXT_TO_SPEECH_REST`

Tool to convert text into natural-sounding speech. Use when you need TTS audio from text inputs.

### Topic Detection

**Slug:** `DEEPGRAM_TOPIC_DETECTION`

Transcribes audio and detects topics discussed in pre-recorded audio files using Deepgram's AI. Use this tool to automatically identify and extract the main topics from audio content, with confidence scores for each detected topic. The tool also provides a full transcript. Requirements: - Audio file must be accessible via a public URL - Only English language is supported for topic detection - Supports common audio formats: WAV, MP3, MPEG, FLAC, WebM Returns detected topics with confidence scores (0.0 to 1.0) and the full audio transcript.