examples

# Deepgram Examples Working examples showing how to use Deepgram with popular platforms, frameworks, and ecosystems. ## When to use examples - You're integrating Deepgram with a specific third-party service (telephony, video, LLM frameworks, bots, cloud platforms) - You want a runnable demo of a complete integration, not just SDK-level code - You want to see Deepgram slotted into a real-world architecture **Use a different skill when:** - You want a minimal feature snippet (one product, one language, < 50 lines) → `recipes` skill - You want a clean starter app to extend with no third-party service → `starters` skill - You want the raw API contract → `api` skill ## Browse examples Repository: <https://github.com/deepgram/examples> Examples are numbered (010, 020, ...) and each is a self-contained integration. ## Category map | Category | Integrations | Common STT choice | |---|---|---| | **Telephony** | Twilio, Vonage, SignalWire, Telnyx, Plivo, Daily.co, Asterisk/FreeSWITCH | Nova live (`/v1/listen`) for call transcription; Flux STT (`/v2/listen`) for AI-agent calls | | **Voice AI frameworks** | LiveKit Agents, Pipecat, OpenAI Agents SDK, CrewAI | Flux STT (`/v2/listen`) — built-in turn detection; or Voice Agent (`/v1/agent/converse`) for full-pipeline | | **Chat platforms** | Discord, Slack, Telegram | Nova prerecorded (`/v1/listen`) for attachments | | **Web frameworks** | Next.js, Nuxt, Django, SvelteKit, NestJS, Express + React, FastAPI, Spring Boot, Gin | Nova live (`/v1/listen`) for captions; Nova prerecorded for batch | | **Mobile / desktop** | Expo, Flutter, Swift iOS, Kotlin Android, Tauri, Electron | Nova live (`/v1/listen`); Flux STT (`/v2/listen`) if the app is a voice agent | | **Cloud / serverless** | AWS Lambda, Cloudflare Workers | Nova prerecorded (`/v1/listen`) — best fit for request/response | | **Meeting platforms** | Zoom, Riverside.fm, Webex (recordings); Jitsi, Microsoft Teams (live meetings) | Nova prerecorded (`/v1/listen`) for recordings; Nova live (`/v1/listen`) for the Jitsi bridge and the Teams bot | | **Browser / no-bundler** | Vanilla JavaScript | Nova live (`/v1/listen`) via `@deepgram/sdk` in the browser | | **LLM frameworks** | LangChain, LlamaIndex, Vercel AI SDK, Haystack, Semantic Kernel | Nova or Flux STT depending on streaming vs batch | | **Low-code / automation** | n8n community nodes | Nova (`/v1/listen`) for event-driven transcription | | **Broadcast** | OBS Studio (native C plugin) | Nova live (`/v1/listen`) for on-screen captions | | **Infrastructure** | Deepgram API proxy servers (`520-node-deepgram-proxy`, `521-deepgram-proxy-python-uv`), Silero VAD segmentation (`530-silero-vad-speech-segmentation-python`), multi-provider LLM proxy for Voice Agent (`530-voice-agent-multi-provider-proxy-python`; two directories share the number 530) | The proxies forward Nova prerecorded, Nova live, and Aura TTS; Silero VAD sends each detected speech segment to Nova prerecorded; the Voice Agent proxy is the agent's `think.endpoint.url`, with STT and TTS staying on the agent | The column above is the STT half of each integration. On the TTS side, Aura (`/v1/speak`) is what these examples use today — **no example in the repo covers Flux TTS (`/v2/speak`) yet**, even in the telephony and voice-AI-framework categories where it fits best. If you're wiring Flux TTS into one of these platforms, take the integration's transport and auth handling from the example and the Flux TTS contract from the `api` skill. ## Install the related SDK skills If your integration uses a specific Deepgram SDK, install its skills for language-idiomatic patterns: ```bash npx skills add deepgram/deepgram-python-sdk # Python npx skills add deepgram/deepgram-js-sdk # Node.js / TypeScript npx skills add deepgram/deepgram-java-sdk # Java npx skills add deepgram/deepgram-go-sdk # Go npx skills add deepgram/deepgram-rust-sdk # Rust npx skills add deepgram/deepgram-dotnet-sdk # C# / .NET ``` Swift and Kotlin SDK skills are not listed because those repositories are not public and `npx skills add` cannot reach them. For browser work, open the `browser-agent` skill: it covers the four Browser Agent SDK packages published on npm (`@deepgram/agents`, `@deepgram/react`, `@deepgram/ui`, `@deepgram/agents-widget`). ## Related Deepgram skills - `api` — consolidated REST + WebSocket API reference - `recipes` — focused feature snippets (one feature, one language) - `starters` — starter apps without third-party integrations - `docs` — documentation finder - `setup-mcp` — Deepgram MCP server installation