> For clean Markdown of any page, append .md to the page URL. > For a complete documentation index, see https://developers.deepgram.com/docs/stt/getting-started/llms.txt. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://developers.deepgram.com/_mcp/server. # Getting Started with Speech to Text With Deepgram’s STT API, you can transcribe both pre-recorded files and real-time streams, choose models optimized for different domains, and integrate transcription directly into your apps for use cases like conversational AI, live captions, and agent assist. Before you start, consider your use case. Deepgram STT offers three main paths: **Streaming audio**
}>
Benefits: Model-integrated end-of-turn detection, configurable turn-taking dynamics
Examples: Contact center agents, customer support bots, real-time assistants.
[Get started](/docs/flux/quickstart)
}>
Benefits: Transcripts in real time, larger language availability, can get diarized transcripts
Examples: Captions, live event transcription, monitoring audio feeds.
[Get started](/docs/live-streaming-audio)
**Pre-recorded audio**
}>
**Benefits:** Simple implementation, broader language availability, cost efficient
**Examples:** Transcribing interviews, podcasts, meetings, support calls.
[Get started](/docs/pre-recorded-audio)