THE COMMUNITY RESOURCE
DIRECTORY
Small decisions.
Small decisions.
Endless possibilities.
Find what people build with Jev. Explore tools, libraries, experiments and ideas from across the TypeSafe ecosystem.
11,066unique resources
29clear categories
220focused subcategories
Yourssave a personal reading list
YOUR NEXT BUILD STARTS HERE
Download catalog
Explore the collection
Browse groups
Voice & real-time interfaces69 of 69 matches
Voice assistants29 of 29 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| Always-on assistant without a wake wordx.com | — | Always-listening voice assistant with no wake word, where Jev decides from probabilities whether speech is a command for the computer or ordinary conversation. | Voice & real-time interfacesVoice assistants | |
| CharaDock Jev fast pathgithub.com | — | Optional Jev fast path in CharaDock, a Windows desktop character companion that talks and works with Codex, deciding music controls from finished utterances and the character's expressions and reactions. | Voice & real-time interfacesVoice assistants | |
| CharaDock Jev fast path — repogithub.com | — | Optional Jev fast path in CharaDock, a Windows desktop character companion that talks and works with Codex, deciding music controls from finished utterances and the character's expressions and reactions. | Voice & real-time interfacesVoice assistants | |
| Cinderpaw Jev voice commandsgithub.com | — | Local desktop AI assistant where Jev can act as the voice provider: speech is transcribed on device or in the cloud and Jev decides which command was meant. | Voice & real-time interfacesVoice assistants | |
| Cinderpaw Jev voice commands — repogithub.com | — | Local desktop AI assistant where Jev can act as the voice provider: speech is transcribed on device or in the cloud and Jev decides which command was meant. | Voice & real-time interfacesVoice assistants | |
| Jamcat voice-driven UIx.com | — | Voice-driven music app where Pipecat hands off in real time between Jev, a command agent that drives the UI, and a voice agent that handles generation and conversation. | Voice & real-time interfacesVoice assistants | |
| Jev voice browser — demox.com | — | Voice control for a real browser that asks 9 to 11 questions per utterance, including a Noul for destructive actions. | Voice & real-time interfacesVoice assistants | |
| Jev voice browser — discussionwww.reddit.com | — | Voice control for a real browser that asks 9 to 11 questions per utterance, including a Noul for destructive actions. | Voice & real-time interfacesVoice assistants | |
| Jev Voice — demox.com | — | Menu-bar Swift app for macOS where you hold a hotkey, speak a command such as "quit spotify" or "set volume to 30", and Jev turns it into a typed macOS action, optionally asking before running. | Voice & real-time interfacesVoice assistants | |
| Jev Voice — sitechris-wozniczek.github.io | — | Menu-bar Swift app for macOS where you hold a hotkey, speak a command such as "quit spotify" or "set volume to 30", and Jev turns it into a typed macOS action, optionally asking before running. | Voice & real-time interfacesVoice assistants | |
| macbrowgithub.com | 123 | Experimental hands-free macOS assistant: speak a command and Jev routes it in about 300 ms to AppleScript tools or a Chrome browser task, with Gradium speech in and out over LiveKit Agents and a safety policy. | Voice & real-time interfacesVoice assistants | |
| macbrow — demowww.youtube.com | — | Experimental hands-free macOS assistant: speak a command and Jev routes it in about 300 ms to AppleScript tools or a Chrome browser task, with Gradium speech in and out over LiveKit Agents and a safety policy. | Voice & real-time interfacesVoice assistants | |
| MacParakeet Voice Controlgithub.com | — | MacParakeet, a local-first voice app for Apple Silicon, adds a Voice Control feature where Jev decides the action to take from a spoken command, gated behind explicit consent. | Voice & real-time interfacesVoice assistants | |
| MacParakeet Voice Control — appmacparakeet.com | — | MacParakeet, a local-first voice app for Apple Silicon, adds a Voice Control feature where Jev decides the action to take from a spoken command, gated behind explicit consent. | Voice & real-time interfacesVoice assistants | |
| Rexclaw Companionsgithub.com | 9 | Self-hosted anime voice companions with 3D VRM avatars on Grok Voice Realtime, where an optional Jev key drives expressive faces, background gestures and group-call turn-taking. | Voice & real-time interfacesVoice assistants | |
| Rexclaw Companions — appcodemarchant.com | — | Self-hosted anime voice companions with 3D VRM avatars on Grok Voice Realtime, where an optional Jev key drives expressive faces, background gestures and group-call turn-taking. | Voice & real-time interfacesVoice assistants | |
| Rexclaw Companions — codegithub.com | 9 | Self-hosted anime voice companions with 3D VRM avatars on Grok Voice Realtime, where an optional Jev key drives expressive faces, background gestures and group-call turn-taking. | Voice & real-time interfacesVoice assistants | |
| Toothlessx.com | — | Always-on voice assistant demo where Jev decides in real time whether what you say is directed at the AI or at someone else, so no wake word is needed. | Voice & real-time interfacesVoice assistants | |
| Voice commands for Autumnx.com | — | Dictation in the Autumn pricing dashboard that turns spoken requests into plan attaches, updates and schedules instead of manual edits. | Voice & real-time interfacesVoice assistants | |
| Voice commands for Fusion 360x.com | — | Voice control for Fusion 360 where you click, speak and Fusion runs the feature, with Jev only deciding whether the utterance is a command. | Voice & real-time interfacesVoice assistants | |
| Voice computer assistantx.com | — | Early computer assistant where local Whisper transcribes everything the author says, and Jev classifies it into actions using the screen's accessibility tree from a small Swift app, all in real time. | Voice & real-time interfacesVoice assistants | |
| Voice control for Mac, honest takex.com | — | Day-long test of controlling a Mac by voice with Jev: actions land in about 700ms and refusals in 350ms, but results depend on well-separated options and it struggles once commands need reasoning. | Voice & real-time interfacesVoice assistants | |
| Voice Cursor Jev demox.com | — | Voice Cursor demo where Jev acts as the live decision layer for spoken edits, such as making an email warmer, keeping Wednesday and dropping the apology. | Voice & real-time interfacesVoice assistants | |
| Voice Cursor Jev demo — appwww.voicecursor.ai | — | Voice Cursor demo where Jev acts as the live decision layer for spoken edits, such as making an email warmer, keeping Wednesday and dropping the apology. | Voice & real-time interfacesVoice assistants | |
| Voice-controlled Chrome with Jevwww.youtube.com | — | Explainer plus a build that drives Chrome by voice through a mic, Whisper, Jev, and Playwright pipeline, set up with the TypeSafe skill in Claude Code, with try-on and news-filter extras. | Voice & real-time interfacesVoice assistants | |
| Voice-controlled Figmax.com | — | Controls Figma by speaking commands that Jev turns into design actions; total cost from first line of code to the demo video was $0.01. | Voice & real-time interfacesVoice assistants | |
| Voice-controlled Mac agentx.com | — | Voice-controlled computer use for the Mac that acts on spoken instructions as they arrive, opening Notes before the user finishes the sentence. | Voice & real-time interfacesVoice assistants | |
| Voice-driven computer usewww.reddit.com | — | Video of spoken commands driving a Mac desktop in near real time, with Jev choosing the computer-use actions. | Voice & real-time interfacesVoice assistants | |
| Youform voice builderx.com | — | Voice form builder in Youform where Jev turns spoken instructions into form-building actions quickly. | Voice & real-time interfacesVoice assistants |
Speech processing4 of 4 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| Jev teleprompterx.com | — | Chinese teleprompter demo where a speech model transcribes the speaker and Jev judges whether they are reading along, speeding up or going off script, then moves the prompt position. | Voice & real-time interfacesSpeech processing | |
| jevisx.com | — | Always-on voice input harness that runs streaming speech-to-text through a Jev directed-speech filter so ambient chatter no longer ends up in AI voice messages. | Voice & real-time interfacesSpeech processing | |
| jevis — repogithub.com | 1 | Always-on voice input harness that runs streaming speech-to-text through a Jev directed-speech filter so ambient chatter no longer ends up in AI voice messages. | Voice & real-time interfacesSpeech processing | |
| Semantic Live Captiongithub.com | 7 | Real-time captioning page that streams speech through the open Confucius4-R2T2 ASR model while Jev annotates each sentence for key points, emotion and intent, with DeepSeek optionally proposing new labels. | Voice & real-time interfacesSpeech processing |
Real-time communication4 of 4 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| JEV Call Screenergithub.com | — | Self-hosted Go call-screening backend that asks callers why they are calling, has Jev classify the transcript, and forwards or rejects the call via a Twilio adapter, failing open on errors or uncertainty. | Voice & real-time interfacesReal-time communication | |
| Speko Gateway Jev voice routergithub.com | — | Open customer-side data plane for real-time voice AI that relays evaluation models such as typesafe:jev-1.13.0, with an example that routes each finalized voice turn using one batched Jev evaluation. | Voice & real-time interfacesReal-time communication | |
| Speko Gateway Jev voice router — appspeko.ai | — | Open customer-side data plane for real-time voice AI that relays evaluation models such as typesafe:jev-1.13.0, with an example that routes each finalized voice turn using one batched Jev evaluation. | Voice & real-time interfacesReal-time communication | |
| Speko Gateway Jev voice router — repogithub.com | — | Open customer-side data plane for real-time voice AI that relays evaluation models such as typesafe:jev-1.13.0, with an example that routes each finalized voice turn using one batched Jev evaluation. | Voice & real-time interfacesReal-time communication |
More projects & source code13 of 13 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| AIAvatarKit Jev turn-end gategithub.com | — | Turn-end gate for AIAvatarKit conversational avatars that asks Jev whether the user is yielding the floor or will keep speaking, adding a semantic signal to silence-based voice activity detection. | Voice & real-time interfacesMore projects & source code | |
| AIAvatarKit turn-end gate — repogithub.com | — | Voice-dialog turn-end detection: 0.5 s after speech stops, Jev gives the probability that the assistant should keep holding the turn, which maps to extra wait time, with Jev taking about 0.22 s. | Voice & real-time interfacesMore projects & source code | |
| aloud Jev judgegithub.com | — | Voice meditation facilitator that asks Jev yes/no questions about each utterance to detect when the speaker has finished and to recognize spoken commands, replacing a Haiku classifier. | Voice & real-time interfacesMore projects & source code | |
| aloud Jev judge — repogithub.com | — | Voice meditation facilitator that asks Jev yes/no questions about each utterance to detect when the speaker has finished and to recognize spoken commands, replacing a Haiku classifier. | Voice & real-time interfacesMore projects & source code | |
| BB Livegithub.com | 8 | Persistent voice channel plugin for the BB workspace built on GPT-Live, with an experimental toggle that uses Jev to classify spoken requests before the intent model. | Voice & real-time interfacesMore projects & source code | |
| BiBi Keyboard Jev classifiergithub.com | — | BiBi Keyboard, an Android LLM+ASR voice input method, uses Jev to pick which prompt to apply to your dictated text after speech recognition. | Voice & real-time interfacesMore projects & source code | |
| Discord Live Voice Botgithub.com | 6 | Discord voice bot that talks back in real time and runs 53 server tools behind a voice-based owner gate, using Jev to judge whether each spoken line was meant for the bot and dropping replies that were not. | Voice & real-time interfacesMore projects & source code | |
| Light-Whispergithub.com | 64 | Windows hold-to-talk dictation app whose Auto modes ask Jev whether AI polish, screen context or web search is needed, keeping the original text when polishing is unnecessary. | Voice & real-time interfacesMore projects & source code | |
| OpenCues decision seamgithub.com | — | Open standard that turns any text field into a two-way LLM channel, using a TypeSafe decision seam to select commands, sentences and candidates, or skip a generation call entirely. | Voice & real-time interfacesMore projects & source code | |
| OpenCues decision seam — docsgithub.com | — | Open standard that turns any text field into a two-way LLM channel, using a TypeSafe decision seam to select commands, sentences and candidates, or skip a generation call entirely. | Voice & real-time interfacesMore projects & source code | |
| Vellum voice judgesgithub.com | — | Jev provider in the Vellum personal assistant used by live-voice judges: one decides whether a barged-in request should keep running in the background, another overrules the fast front door when a turn needs tools. | Voice & real-time interfacesMore projects & source code | |
| VoiceX llm-bench Jev judgegithub.com | — | Benchmark tool for the VoiceX desktop voice-input app that uses Jev as the judge of LLM post-processing, asking three fixed questions about each ASR correction to credit good edits and flag bad ones. | Voice & real-time interfacesMore projects & source code | |
| VoiceX llm-bench Jev judge — repogithub.com | — | Benchmark tool for the VoiceX desktop voice-input app that uses Jev as the judge of LLM post-processing, asking three fixed questions about each ASR correction to credit good edits and flag bad ones. | Voice & real-time interfacesMore projects & source code |
More posts & discussions13 of 13 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| AIAvatarKit turn-end gatex.com | — | Voice-dialog turn-end detection: 0.5 s after speech stops, Jev gives the probability that the assistant should keep holding the turn, which maps to extra wait time, with Jev taking about 0.22 s. | Voice & real-time interfacesMore posts & discussions | |
| Automatic IME mode switchingx.com | — | Prototype that watches typing, uses Jev to judge Japanese versus English, switches the keyboard IME mode automatically, and converts already-typed romaji like "watashi" into Japanese. | Voice & real-time interfacesMore posts & discussions | |
| Designing at the speed of voicex.com | — | Speaks UI layouts into existence: a local transcription model feeds Jev, which assembles shadcn components into the described layout. | Voice & real-time interfacesMore posts & discussions | |
| Directed-speech filter with Deepgram Fluxx.com | — | Voice setup where Deepgram Flux transcribes everything said in the room and Jev answers one question per utterance: whether any of it was meant for the assistant. | Voice & real-time interfacesMore posts & discussions | |
| Emoji Jev — demox.com | — | Emoji autocomplete that lights up a keyboard of up to 254 emojis as you type; each pause is one Jev call with eight parallel questions (emoji, emotion, mood, urgency, energy, sarcasm, joke, wants reply). | Voice & real-time interfacesMore posts & discussions | |
| Emoji Jev — discussionwww.reddit.com | — | Emoji autocomplete that lights up a keyboard of up to 254 emojis as you type; each pause is one Jev call with eight parallel questions (emoji, emotion, mood, urgency, energy, sarcasm, joke, wants reply). | Voice & real-time interfacesMore posts & discussions | |
| Instant spreadsheet checksx.com | — | Collaborative AI spreadsheet demo where Jev runs 20+ checks every time a cell is edited and shows warnings immediately, with no LLM in the loop. | Voice & real-time interfacesMore posts & discussions | |
| Jev + Rime input methodx.com | — | Test pairing Jev with the open-source Rime (Squirrel) Chinese input method to rank the wanted character higher among candidates while typing. | Voice & real-time interfacesMore posts & discussions | |
| jev boardx.com | — | Experimental keyboard that uses Jev for ambient, as-you-type intelligence in the interface, an exploration at the edges of generative UI. | Voice & real-time interfacesMore posts & discussions | |
| Live slide selection while speakingx.com | — | Presentation demo where Jev decides in real time which slide to pull up based on what the speaker is saying. | Voice & real-time interfacesMore posts & discussions | |
| Point-and-speak canvasx.com | — | Whiteboard canvas controlled by pointing at a webcam and speaking, where Jev picks the action, target and location; the post shares only a video, with no code or data, and one frame shows a 262 ms judgment. | Voice & real-time interfacesMore posts & discussions | |
| Real-time voice browser controlx.com | — | Voice control for the browser: the spoken transcript goes to Jev, which returns probabilities in ~300ms and triggers the click, at $0.0002 per decision. | Voice & real-time interfacesMore posts & discussions | |
| Voice intent to tool callsx.com | — | Voice-agent experiment that replaces the LLM's tool-choice step with Jev so decisions run on partial transcripts, letting the agent act before the user finishes the sentence. | Voice & real-time interfacesMore posts & discussions |
More guides & websites6 of 6 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| aloud Jev judge — appaloud.rest | — | Voice meditation facilitator that asks Jev yes/no questions about each utterance to detect when the speaker has finished and to recognize spoken commands, replacing a Haiku classifier. | Voice & real-time interfacesMore guides & websites | |
| BiBi Keyboard Jev classifier — appbibi.brycewg.com | — | BiBi Keyboard, an Android LLM+ASR voice input method, uses Jev to pick which prompt to apply to your dictated text after speech recognition. | Voice & real-time interfacesMore guides & websites | |
| Emoji Jev — appemoji-jev.whop.site | — | Emoji autocomplete that lights up a keyboard of up to 254 emojis as you type; each pause is one Jev call with eight parallel questions (emoji, emotion, mood, urgency, energy, sarcasm, joke, wants reply). | Voice & real-time interfacesMore guides & websites | |
| IntelliPrompter — appfinetuningsingh.github.io | — | Teleprompter of talking points that checks each point off once you have covered it in your own words; running on Jev costs at most about $0.50 an hour versus about $40 on LLMs. | Voice & real-time interfacesMore guides & websites | |
| OpenCues decision seam — appwww.opencues.com | — | Open standard that turns any text field into a two-way LLM channel, using a TypeSafe decision seam to select commands, sentences and candidates, or skip a generation call entirely. | Voice & real-time interfacesMore guides & websites | |
| Vellum voice judges — appvellum.ai | — | Jev provider in the Vellum personal assistant used by live-voice judges: one decides whether a barged-in request should keep running in the background, another overrules the fast front door when a turn needs tools. | Voice & real-time interfacesMore guides & websites |
No matches this time
Try a broader search or reset your filters to explore the full collection.
Showing 69 of 69 resources