Things to build with on-device models

Review your app and find features you could move on the device.

Add a private, offline AI feature to an app that has none.

Build a full podcast studio: transcribe, clip, and title on the device.

Record, clean, transcribe, and clip: a whole creator pipeline, offline.

Build a voice-notes app that cleans the audio and transcribes it offline.

Detect the language of a voice note, then transcribe with the right model.

Transcribe interviews and auto-redact names before you share them.

Detect the language, then redact PII in that language.

A clean, filler-free transcript from a raw, noisy recording.

Tag a post and suggest an emoji for it.

Turn a webinar into teaser clips for social.

Build a fully local clipping engine: transcribe, clip, title, caption.

Build an OpenClaw skill that tags and language-routes text.

A CLI that clips the highlights out of a video.

Build a Mac clipping app: drop a video, get ranked shorts.

A Raycast command that redacts or tags the clipboard.

A chat bot that tags and language-routes messages, in Node.

Build an agent skill that tags and language-routes text for an agent.

Build an OpenClaw skill that pulls the best clips out of a video.

Build a Claude Code skill that clips highlights from a video.

A Goose extension that transcribes and clips media, locally.

Give Open Interpreter on-device transcription and PII redaction.

Build a podcast player that transcribes every episode and makes it searchable, on the device.

Build a meeting recorder that hands you a timestamped transcript before you leave the room.

Turn voice memos into searchable, editable text notes, offline.

Generate SRT subtitles for any video file on the Mac.

Add captions to a video player that run on the device.

Build an interview app for journalists where sources never leave the phone.

Highlight each word as it's spoken, timed to the audio.

Tighten a system transcriber's word timings so captions land on the word.

Build a text-based video editor where selecting words trims the clip.

Add bouncing word-by-word captions to vertical videos.

Make transcript search jump the audio to the exact word.

Word-perfect captions from any recording.

Add a Shortcuts action that transcribes a voice memo on the device.

A CLI that transcribes a folder of recordings to text and SRT.

A Final Cut helper that generates word-accurate captions.

Build an OpenClaw skill that transcribes any audio or video you point it at.

Build a Claude and Codex skill that transcribes audio, locally.

Add a one-tap 'clean audio' button that removes noise and room echo.

Master a podcast episode to platform loudness, on the device.

Rescue a recording made in a noisy, untreated room.

Clean the audio on a phone video in one pass.

Add a 'remove ums' button to a podcast or video editor.

Build a speaking coach that counts and marks your filler words.

Show a filler-word timeline for any recording.

Trim every filler from a talking-head take automatically.

A Shortcut that cleans up any audio file's sound.

A CLI that batch-enhances a podcast back catalog.

Build a Mac audio-cleanup app: drag in a recording, get studio sound.

Turn any recording into a ranked set of shorts.

Generate a highlight reel from a long stream or talk.

Produce a rough auto-edit from a rambling take.

Clip a whole back catalog of videos overnight on a Mac.

Write an MCP tool that strips personal data before it reaches any other tool.

Scrub personal data out of any text before it hits an LLM.

Redact PII from logs on the device before they're stored.

Run PII redaction in a Lambda so data never leaves your infrastructure.

Scrub a support form before it submits.

De-identify a dataset before fine-tuning or sharing it.

Build an OpenClaw skill that redacts PII before any tool sees it.

A Shortcut that redacts PII from the clipboard.

A CLI that redacts PII from every file in a directory.

A VS Code extension that strips PII and secrets from a file.

Build an agent skill that redacts PII before your coding agent sends anything out.

Build a Codex skill that scrubs PII and secrets before a commit.

An OpenHands tool that de-identifies data before the agent touches it.

An OpenClaw skill that redacts a message before it hits Slack or WhatsApp.

Detect a caller's language and route the voicemail.

Sort an archive of recordings by spoken language.

Route incoming messages to the right language agent.

Auto-switch the keyboard and spellcheck to the language being typed.

Tag user posts by language for search and moderation.

Pick the dominant language of a recording to route it.

Detect the language of a voicemail or dictation before transcribing.

Gate transcription so it only runs on languages you support.

Detect the language of a search query and pick the right index.

Auto-tag articles and posts with their topics for a feed.

Route content to channels by topic, no server call.

Organize a reading list by topic, locally.

Give a recommender a topic-per-item signal that runs on the phone.

Auto-name and describe every clip and file.

Title a note the moment it's saved.

Write library descriptions for a media archive.

Draft an SEO title and meta description from a page's text.

Auto-file support tickets into the right queue by topic.

A Shortcut that titles and describes whatever you share to it.

An Obsidian plugin that tags notes with their topics, on the device.

Build an agent skill that titles and describes any file the agent makes.

Add an emoji suggestion strip above the keyboard.

Suggest a reaction emoji for any message.

Auto-add an emoji to a task or calendar title.

Build an emoji search that understands a sentence, not just keywords.

Suggest a status emoji from what someone is typing.

Build a whiteboard that snaps rough sketches into clean shapes.

Clean up a freehand diagram into crisp shapes.

Add shape recognition to a web canvas with no download.

Snap flowchart boxes and connector lines as you sketch them.

Add smart-shape drawing to a kids' learning app.

Recognize gesture shapes to trigger app actions.