Crate deepgram

Expand description

Official Rust SDK for Deepgram’s automated speech recognition APIs.

Get started transcribing with a Transcription object.

§Cargo features

listen (default): speech-to-text, REST and WebSocket, including Flux.
speak (default): text-to-speech, REST and WebSocket, including Flux.
manage (default): project, key, and usage management.
read (default): Text Intelligence over /v1/read; see the read module.
agent (default): the Voice Agent WebSocket client; see the agent module.
connect-diagnostics: per-phase connect timings for /v1/listen WebSocket connections; see the diagnostics module.
rustls-tls-native-roots: also trust the operating system’s certificate store for wss:// WebSocket connections, on top of the bundled public roots. For TLS-inspecting proxies, internal CAs, and self-hosted deployments; see the tls module.

agent: Voice Agent (/v1/agent/converse) types.
auth: Token-based authentication for Deepgram.
common: Common lib for other modules
diagnostics: Opt-in connect-time diagnostics for live transcription streaming connections (/v1/listen).
listen: Listen module
manage: Manage module
read: Analyze text with Deepgram’s Text Intelligence (/v1/read) API.
speak: Speak module
tls: TLS trust for wss:// WebSocket connections.

Deepgram: A client for the Deepgram API.
HttpError: A generic “error” for HTTP connections
ReqwestError: The Errors that may occur when processing a Request.
SerdeJsonError: This type represents all possible errors that can occur when serializing or deserializing JSON data.
Speak: Generate speech from text using Deepgram’s text to speech api.
Transcription: Transcribe audio using Deepgram’s automated speech recognition.

DeepgramError: Errors that may arise from the deepgram crate.
SerdeUrlencodedError: Errors returned during serializing to application/x-www-form-urlencoded.
TungsteniteError: Possible WebSocket errors.