Skills

Skill 7 of 14

Recipes

Find focused, runnable Deepgram recipes for a specific feature × language.

2 minutes · 501 words · 8 sections

Install

$ npx skills add deepgram/skills --skill recipes

$ npx skills add deepgram/skills

» /plugin marketplace add deepgram/skills

The first command installs just this skill, by the name in its SKILL.md; the second installs the whole repository.

Deepgram Recipes

Agent-maintained micro-recipes showing how to use every Deepgram SDK feature across every supported language. Each recipe is a focused, runnable snippet — not a full app.

When to use recipes

You know the product and feature; you want the shortest working code
You want example.py / example.js / example.go / etc. you can copy into your project
You want a language-specific answer to “how do I call {feature} with the Deepgram SDK?”

Use a different skill when:

You want a full starter app with a web UI, deploy config, etc. → starters skill
You want integration with a third-party platform (Twilio, LiveKit, Vercel AI SDK, Discord, etc.) → examples skill
You want the full API contract (params, responses, message shapes) → api skill
You want a shell command rather than application code → cli skill

Browse recipes

Repository: https://github.com/deepgram/recipes (opens in a new tab)

Coverage matrix: https://github.com/deepgram/recipes/blob/main/COVERAGE.md (opens in a new tab)

Recipe structure

recipes/{language}/{product}/{version}/{recipe}/
  example.{ext}       # runnable, < 50 lines, reads DEEPGRAM_API_KEY from env
  example_test.{ext}  # runs the example as a subprocess, asserts output
  README.md           # feature explanation, params, sample output, how to run

Products covered

Product	Recipe examples
Speech-to-Text — Nova (`/v1/listen`)	transcribe-url, transcribe-file, paragraphs, diarize, smart-format, utterances, summarize, sentiment, topics, intents, detect-entities, detect-language, redact, search, keywords, streaming
Speech-to-Text — Flux STT (`/v2/listen`)	streaming conversational transcription, EOT / eager-EOT, mid-session `Configure`, keyterms
Text-to-Speech — Aura (`/v1/speak`)	generate-audio, stream-audio, websocket-streaming, select-model, select-encoding, bit-rate
Audio Intelligence (`/v1/listen`)	summarize, sentiment, topics, intents, entities
Voice Agents	connect, custom-llm, custom-tts, function-calling
Text Analysis (`/v1/read`)	summarize, sentiment, topics, intents

Audio Intelligence and Text Analysis run the same analysis on different inputs, so pick the row by what you already have. The Audio Intelligence recipes are query parameters layered on /v1/listen, so the input is audio and entities is one of them. The Text Analysis recipes are one POST /v1/read, so the input is text and there is no entities recipe: detect_entities on /v1/read returns 400 {"err_code":"INVALID_QUERY_PARAMETER","err_msg":"unknown query parameter: detect_entities"}. Open the audio-intelligence skill for the first and text-intelligence for the second. Text Analysis recipes exist in all seven languages.

Nova is the general-purpose STT family; Flux STT is designed for conversational audio and voice agents. Both are actively maintained — see the api skill’s “Nova vs Flux STT” section for the decision guide. Note that Flux STT (/v2/listen) and Flux TTS (/v2/speak) are separate products that share the Flux name.

Flux TTS (/v2/speak) has no recipes yet. The TTS recipes above cover Aura on /v1/speak only — there is no text-to-speech/v2 directory in the recipes repo. For Flux TTS, use the api skill’s “Aura vs Flux TTS” section for the endpoint contract, or the starters skill’s flux-tts apps (node, flask, fastapi, django, java) for runnable code.

Languages

Python, JavaScript, Go, .NET, Java, Rust, plus the Deepgram CLI (dg / deepctl).

For language-idiomatic patterns beyond a single recipe (full quick-starts, common patterns, gotchas), install the SDK-specific skills:

npx skills add deepgram/deepgram-python-sdk     # Python
npx skills add deepgram/deepgram-js-sdk         # JavaScript / 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).

api — consolidated REST + WebSocket API reference
examples — third-party platform integrations (Twilio, LiveKit, LangChain, etc.)
starters — runnable starter apps (framework × feature matrix)
docs — documentation finder
setup-mcp — Deepgram MCP server installation

About this skill

Trigger

Find focused, runnable Deepgram recipes for a specific feature × language. Use whenever someone wants a minimal working code snippet for ONE feature (transcribe URL, diarize, smart-format, voice agent connect, etc.) rather than a full starter app. Recipes are under 50 lines, read DEEPGRAM_API_KEY from env, and ship with a runnable example_test. Covers Python, JavaScript, Go, .NET, Java, Rust, and the Deepgram CLI.

The verbatim description from this skill’s front matter — the string an agent matches on to decide whether to load it.

Colophon for this issue

Colophon

How this issue was made

Source

github.com/deepgram/skills, branch main, last pushed 20 September 2026.

Licence

Discovery

14 skills found by walking the repository tree for SKILL.md, not by matching a directory convention. One layout observed: skills/*/SKILL.md.

Issue colours

Resolved from a deterministic hash of the owner name. Two accent tones are generated per issue and each is proven against its own ground before it ships: a single accent that passes AA on both light and dark paper is arithmetically impossible.

Heading repairs

1 repair applied to this skill so the document has one h1 and no skipped levels:

Shifted “8 headings” from h1 to h2 so the skill title is the only h1.

Spec compliance

Every skill satisfies the Agent Skills specification’s frontmatter rules.

Images

Images inside a skill come from the upstream repository. Where the author gave no alternative text we mark the image decorative rather than inventing a description — a plausible caption we made up is worse than none for the reader who depends on it.

Marketplace

A plugin manifest is published at .claude-plugin/marketplace.json by Deepgram Developer Relations, declaring 7 plugins. It is read for editorial metadata only — never as the skill index, which is always the repository tree.

Signal

Install counts come from skills.sh. They measure downloads, not quality, and an unranked repository is not an unread one.

Agent surfaces

The whole issue is available as one markdown document at /deepgram/skills.md, and each skill at its own .md URL.

Publication

Set by Skills Docs from the source repository. Body text is Literata at the reader’s chosen size and measure; code is Geist Mono. Nothing on this page was written by us except this paragraph.

Deepgram Recipes

Agent-maintained micro-recipes showing how to use every Deepgram SDK feature across every supported language. Each recipe is a focused, runnable snippet — not a full app.

You know the product and feature; you want the shortest working code

You want example.py / example.js / example.go / etc. you can copy into your project

You want a language-specific answer to “how do I call {feature} with the Deepgram SDK?”

Use a different skill when:

You want a full starter app with a web UI, deploy config, etc. → starters skill

You want integration with a third-party platform (Twilio, LiveKit, Vercel AI SDK, Discord, etc.) → examples skill

You want the full API contract (params, responses, message shapes) → api skill

You want a shell command rather than application code → cli skill

recipes/{language}/{product}/{version}/{recipe}/
  example.{ext}       # runnable, < 50 lines, reads DEEPGRAM_API_KEY from env
  example_test.{ext}  # runs the example as a subprocess, asserts output
  README.md           # feature explanation, params, sample output, how to run

Product	Recipe examples
Speech-to-Text — Nova (`/v1/listen`)	transcribe-url, transcribe-file, paragraphs, diarize, smart-format, utterances, summarize, sentiment, topics, intents, detect-entities, detect-language, redact, search, keywords, streaming
Speech-to-Text — Flux STT (`/v2/listen`)	streaming conversational transcription, EOT / eager-EOT, mid-session `Configure`, keyterms
Text-to-Speech — Aura (`/v1/speak`)	generate-audio, stream-audio, websocket-streaming, select-model, select-encoding, bit-rate
Audio Intelligence (`/v1/listen`)	summarize, sentiment, topics, intents, entities
Voice Agents	connect, custom-llm, custom-tts, function-calling
Text Analysis (`/v1/read`)	summarize, sentiment, topics, intents

Python, JavaScript, Go, .NET, Java, Rust, plus the Deepgram CLI (dg / deepctl).

For language-idiomatic patterns beyond a single recipe (full quick-starts, common patterns, gotchas), install the SDK-specific skills:

npx skills add deepgram/deepgram-python-sdk     # Python
npx skills add deepgram/deepgram-js-sdk         # JavaScript / 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

api — consolidated REST + WebSocket API reference

examples — third-party platform integrations (Twilio, LiveKit, LangChain, etc.)

starters — runnable starter apps (framework × feature matrix)

docs — documentation finder

setup-mcp — Deepgram MCP server installation