mcpField
E
examples

examples

github.com/deepgramDevelopment26316 installsUpdated 4w ago
Repository

About examples

Deepgram integration examples for developers building with third-party platforms and frameworks. Each numbered example is a self-contained runnable integration, covering speech recognition and speech synthesis patterns.

Key features

  1. 01Runnable, self-contained integrations
  2. 02Examples grouped by ecosystem
  3. 03Speech-to-text integration patterns
  4. 04Aura text-to-speech examples
  5. 05Guidance for related Deepgram skills

Use cases

  1. 01Integrating Deepgram with telephony services
  2. 02Adding transcription to chat or meeting apps
  3. 03Using Deepgram in web or mobile frameworks
  4. 04Building voice AI framework integrations

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/deepgram/skills --skill examples

After installing, type/examplesin Claude Code to confirm it loaded.

Details

From the project README

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

CategoryIntegrationsCommon STT choice
TelephonyTwilio, Vonage, SignalWire, Telnyx, Plivo, Daily.co, Asterisk/FreeSWITCHNova live (/v1/listen) for call transcription; Flux STT (/v2/listen) for AI-agent calls
Voice AI frameworksLiveKit Agents, Pipecat, OpenAI Agents SDK, CrewAIFlux STT (/v2/listen) — built-in turn detection; or Voice Agent (/v1/agent/converse) for full-pipeline
Chat platformsDiscord, Slack, TelegramNova prerecorded (/v1/listen) for attachments
Web frameworksNext.js, Nuxt, Django, SvelteKit, NestJS, Express + React, FastAPI, Spring Boot, GinNova live (/v1/listen) for captions; Nova prerecorded for batch
Mobile / desktopExpo, Flutter, Swift iOS, Kotlin Android, Tauri, ElectronNova live (/v1/listen); Flux STT (/v2/listen) if the app is a voice agent
Cloud / serverlessAWS Lambda, Cloudflare WorkersNova prerecorded (/v1/listen) — best fit for request/response
Meeting platformsZoom, 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-bundlerVanilla JavaScriptNova live (/v1/listen) via @deepgram/sdk in the browser
LLM frameworksLangChain, LlamaIndex, Vercel AI SDK, Haystack, Semantic KernelNova or Flux STT depending on streaming vs batch
Low-code / automationn8n community nodesNova (/v1/listen) for event-driven transcription
BroadcastOBS Studio (native C plugin)Nova live (/v1/listen) for on-screen captions
InfrastructureDeepgram 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:

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

Comments

Markdown supported