verba

KO | EN

gitlite — search

#agent #pytorch #gpt #rag #video #golang #tailwindcss #self-hosted #animation #large-language-models

(★ 9)

Learn a language by talking to an AI that never gets bored, never judges, and — if you want — never touches the internet.

File Explorer

ZIP Download

- launch.json
- - check.yml
  - release.yml
- .gitignore
- project.yml
- extensions.json
- - 1-verba-onboarding-spec.md
  - 2-verba-ana-ekran-ve-ayarlar-spec.md
  - 3-verba-activity-layer-spec.md
  - 4-verba-repair-katmani-spec.md
  - 5-verba-akicilik-ve-monitor-katmani-spec.md
  - 6-verba-erisilebilirlik-spec.md
  - 99-updates.md
  - M4-README.md
  - M5-README.md
  - M6-README.md
  - PLAN-001-setup-screens-0-2.md
  - PLAN-001.done.md
  - PLAN-002-setup-screen-3-model.md
  - PLAN-002.done.md
  - PLAN-003-setup-screen-4-level.md
  - PLAN-003.done.md
  - PLAN-004-setup-screen-5-day-one.md
  - PLAN-004.done.md
  - PLAN-005-setup-rules.md
  - PLAN-005.done.md
  - PLAN-006-memory-scheduler.md
  - PLAN-006.done.md
  - PLAN-007-memory-card-origin.md
  - PLAN-007.done.md
  - PLAN-008-memory-deck-view.md
  - PLAN-008.done.md
  - PLAN-009-signals-carry-measurement.md
  - PLAN-009.done.md
  - PLAN-010-coach-metrics.md
  - PLAN-010.done.md
  - PLAN-011-coach-honesty.md
  - PLAN-011.done.md
  - PLAN-012-plan-dependencies.md
  - PLAN-012.done.md
  - PLAN-013-today-four-states.md
  - PLAN-013.done.md
  - PLAN-014-weakness-to-tomorrow.md
  - PLAN-014.done.md
  - PLAN-015-house-format.md
  - PLAN-016-four-states.md
  - PLAN-017-talk-catalogue-persona-goals.md
  - PLAN-018-talk-voice-primary.md
  - PLAN-019-talk-confidence.md
  - PLAN-020-talk-reflection-history.md
  - PLAN-021-talk-subtitles.md
  - PLAN-022-read-passage-contract.md
  - PLAN-023-read-note-contract.md
  - PLAN-024-read-close-and-prompter.md
  - PLAN-025-listen-timeline.md
  - PLAN-026-listen-transcript-and-answers.md
  - PLAN-027-repair-inventory.md
  - PLAN-028-baseline-and-breakdown-signals.md
  - PLAN-029-bluff-decision.md
  - PLAN-030-rewind.md
  - PLAN-031-controlled-difficulty.md
  - PLAN-032-patience-and-praise.md
  - PLAN-033-memory-in-conversation.md
  - PLAN-034-rehearsal-mode.md
  - PLAN-034-review.md
  - PLAN-035-brought-content.md
  - PLAN-036-listening-conditions.md
  - PLAN-037-surfaces.md
  - PLAN-038-three-defects.md
- - speech.png
  - talk.png
  - today.png
- architecture.md
- demo-poster.jpg
- social-preview.png
- logo.png
- logo.svg
- .thumbnail
- support.js
- Verba - AI Language Learning.dc.html
- check.mjs
- manifest.mjs
- version.mjs
- - react.svg
- - backup.check.ts
  - backup.ts
  - breakdown.check.ts
  - breakdown.ts
  - brought.check.ts
  - brought.ts
  - bundled.ts
  - choices.check.ts
  - choices.ts
  - coach.ts
  - coachmetrics.check.ts
  - coachmetrics.ts
  - conditions.check.ts
  - conditions.loader.mjs
  - conditions.mock-db.mjs
  - conditions.mock-providers.mjs
  - conditions.mock-speech.mjs
  - conditions.ts
  - confidence.check.ts
  - confidence.ts
  - corrections.check.ts
  - db.ts
  - deck.check.ts
  - deck.ts
  - difficulty.check.ts
  - difficulty.ts
  - fmt.check.ts
  - fmt.ts
  - invariants.check.ts
  - keys.check.ts
  - keys.ts
  - lang.check.ts
  - langs.ts
  - learn.check.ts
  - learn.ts
  - level-gap-gate.check.ts
  - level.ts
  - listening.check.ts
  - listening.ts
  - memory.check.ts
  - metrics.ts
  - model.check.ts
  - model.ts
  - models.check.ts
  - models.ts
  - notes.check.ts
  - notes.ts
  - onboarding.check.ts
  - passage.check.ts
  - passage.ts
  - patience.check.ts
  - patience.ts
  - phase2.check.ts
  - phase3.check.ts
  - phase4.check.ts
  - placement.check.ts
  - placement.ts
  - prompter.check.ts
  - prompter.ts
  - prompts.check.ts
  - prompts.ts
  - questions.check.ts
  - questions.ts
  - readaloud.check.ts
  - readaloud.ts
  - reading.check.ts
  - reading.ts
  - rehearsal.check.ts
  - rehearsal.loader.mjs
  - rehearsal.mock-db.mjs
  - rehearsal.mock-providers.mjs
  - rehearsal.mock-speech.mjs
  - rehearsal.ts
  - repair.check.ts
  - repair.ts
  - rewind.check.ts
  - rewind.ts
  - rules.check.ts
  - rules.ts
  - scenarios.check.ts
  - scenarios.ts
  - settings.check.ts
  - settings.ts
  - settingsIndex.check.ts
  - settingsIndex.ts
  - setup.check.ts
  - signals.check.ts
  - signals.ts
  - speech.check.ts
  - speech.ts
  - srs.check.ts
  - srs.ts
  - states.check.ts
  - surfaces.check.ts
  - surfaces.ts
  - text.ts
  - timeline.check.ts
  - timeline.ts
  - update.ts
  - useDay.ts
  - useListening.ts
  - useRead.ts
  - useTalk.ts
  - vault.ts
  - vocab.check.ts
  - vocab.ts
  - voice.check.ts
  - voice.ts
  - voices.check.ts
  - weakness.check.ts
  - weakness.ts
- - Coach.tsx
  - DataPanel.tsx
  - Hints.tsx
  - Listening.tsx
  - Memory.tsx
  - Onboarding.tsx
  - QuestionCard.tsx
  - Read.tsx
  - Settings.tsx
  - States.tsx
  - Talk.tsx
  - Today.tsx
- App.tsx
- main.tsx
- theme.css
- vite-env.d.ts
- - default.json
- - 128x128.png
  - [email protected]
  - 32x32.png
  - 64x64.png
  - icon.icns
  - icon.ico
  - icon.png
  - make-source.py
  - source.png
  - Square107x107Logo.png
  - Square142x142Logo.png
  - Square150x150Logo.png
  - Square284x284Logo.png
  - Square30x30Logo.png
  - Square310x310Logo.png
  - Square44x44Logo.png
  - Square71x71Logo.png
  - Square89x89Logo.png
  - StoreLogo.png
- - lib.rs
  - main.rs
  - speech.rs
  - update.rs
  - vault.rs
- .gitignore
- build.rs
- Cargo.lock
- Cargo.toml
- Info.plist
- tauri.conf.json
- - demo-poster.jpg
  - demo.mp4
  - logo.png
  - logo.svg
  - og.png
  - speech.png
  - talk.png
  - today.png
- - index.html
- - index.html
- - index.html
- - index.html
- - index.html
- - index.html
- - index.html
- - de.json
  - en.json
  - es.json
  - fr.json
  - it.json
  - ja.json
  - pt.json
  - tr.json
- - index.html
- _headers
- _redirects
- build.cjs
- index.html
- README.md
- robots.txt
- sitemap.xml
- styles.css
- template.html
.gitignore
CONTRIBUTING.md
index.html
LICENSE
package-lock.json
package.json
README.md
tsconfig.json
tsconfig.node.json
vite.config.ts

# Use via CDN

jsDelivr

jsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.

Version

File

Link

Example

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

**Learn a language by talking to an AI that never gets bored, never judges,
and — if you want — never touches the internet.**

Full conversations with inline corrections, AI-generated graded reading, automatic
vocabulary capture, and a voice on both ends. Runs entirely on your machine.

[![Download](https://img.shields.io/github/v/release/nuvocode/verba?style=for-the-badge&label=Download&color=2563eb)](https://github.com/nuvocode/verba/releases/latest)
[![License](https://img.shields.io/badge/license-MIT-black?style=for-the-badge)](https://raw.githubusercontent.com/nuvocode/verba/master/LICENSE)
[![Platforms](https://img.shields.io/badge/macOS%20·%20Windows%20·%20Linux-grey?style=for-the-badge)](https://raw.githubusercontent.com/nuvocode/verba/master/)

<sub><a href="https://youtu.be/5JQC3_omELw">▶︎ Click to play — one session end to end: talk, get corrected, hear the reply.</a></sub>

</div>

---

![Today — the daily session plan](https://raw.githubusercontent.com/nuvocode/verba/master/docs/screenshots/today.png)

**A session is a plan, not a menu.** One theme threads today's conversation,
reading, role-play, and the words that are due.

![Talk — conversation with the coach](https://raw.githubusercontent.com/nuvocode/verba/master/docs/screenshots/talk.png)

**The coach speaks first and keeps the turn moving.** Suggestions are there for the
moment you stall — not to answer for you.

![Settings → Speech — the bundled tier](https://raw.githubusercontent.com/nuvocode/verba/master/docs/screenshots/speech.png)

**Voice and dictation, no setup.** A 21 MB voice running inside the app: no key, no
server, nothing on the network.

---

## Why Verba

- **Offline, genuinely.** Point it at [Ollama](https://ollama.com) and the whole app
  — chat, reading, speech, dictation — works on a plane. No account, ever.
- **Your provider, your key.** Ollama, LM Studio, OpenAI, Anthropic, Gemini,
  OpenRouter. Swap in Settings; nothing is locked in.
- **It talks back.** Speak, get transcribed, get answered out loud. The bundled
  models download on demand and run in-process.
- **It remembers.** Words you stumble on become a spaced-repetition deck, and the
  deck feeds tomorrow's session.
- **Your data stays put — and it's yours to move.** Everything lives in a local
  SQLite file. There is no server to sign up to, because there is no server. Export
  it to one file whenever you like, or point Verba at a folder your own sync service
  watches (iCloud Drive, Google Drive, Syncthing, a USB stick) and pick up on another
  machine — same words, same history, nothing to set up again.
- **Eight languages, and adding one is a folder.** English, Spanish, French, German,
  Japanese, Turkish, Italian, Portuguese — see [CONTRIBUTING.md](./CONTRIBUTING.md).

## Download

**[→ Latest release](https://github.com/nuvocode/verba/releases/latest)**

| Platform | | |
|---|---|---|
| **macOS** — Apple Silicon (M1–M4) | [`.dmg`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba_0.6.0_aarch64.dmg) | |
| **macOS** — Intel | [`.dmg`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba_0.6.0_x64.dmg) | |
| **Windows** | [`.exe`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba_0.6.0_x64-setup.exe) | [`.msi`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba_0.6.0_x64_en-US.msi) |
| **Linux** | [`.AppImage`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba_0.6.0_amd64.AppImage) | [`.deb`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba_0.6.0_amd64.deb) · [`.rpm`](https://github.com/nuvocode/verba/releases/download/v0.6.0/Verba-0.6.0-1.x86_64.rpm) |

On a Mac, take **aarch64** for Apple Silicon and **x64** only for an Intel Mac — the
wrong one either refuses to start or crawls under Rosetta.

<details>
<summary><b>The build is unsigned — here's how to open it anyway</b></summary>

<br>

Verba has no Apple Developer certificate and no Windows code-signing certificate, and
we would rather ship the build than gate it behind either. So both operating systems
will refuse to open it on the first try.

**macOS** says *"Apple could not verify 'Verba' is free of malware"* and offers you
**Move to Trash**. That is not a detection — nothing was scanned and nothing was
found. It is what macOS says about *any* app that hasn't been through Apple's
notarisation service. Drag Verba to `/Applications`, then:

```bash
xattr -dr com.apple.quarantine /Applications/Verba.app
```

Without a terminal: try to open it once, then **System Settings → Privacy & Security**
→ scroll down → **Open Anyway**. (On macOS 15+ the old right-click → *Open* trick no
longer works.)

**Windows** SmartScreen shows *"Windows protected your PC"* → **More info** → **Run
anyway**.

Building it yourself sidesteps both, and is the honest option if the above makes you
uncomfortable. None of this is unique to Verba — it's the toll every unsigned
open-source desktop app pays.

</details>

## First run

1. Pick a **language pack** and a **provider** in Settings. `Test connection` lists
   your installed Ollama models.
2. Optional — **Settings → Speech → Bundled models** downloads a voice and a
   transcriber that run inside the app:

| | Model | Size | Speed |
   |---|---|---|---|
   | Voice | **Piper** — one voice per language | ~21 MB | ~7× real time |
   | Voice | **Kokoro** — 53 voices, six languages | 349 MB | ~2× real time |
   | Dictation | **Whisper** base / small | 208 / 639 MB | faster than real time |

Piper is the default a pack recommends — a fiftieth of the download and several
   times quicker, which for a conversation turn matters more than the last few
   percent of naturalness.

3. Hit 🎤 in Talk and say something.

Already run your own speech server, or prefer a cloud voice? Any OpenAI-compatible
server works (Kokoro-FastAPI, speaches, LocalAI, vLLM), as do ElevenLabs and Deepgram
keys. Verba walks **bundled → local server → cloud key → OS voice** and always tells
you which tier is actually speaking.

> **Platform honesty:** all four targets are packaged from the same source by CI, but
> the bundled speech engine has only ever *run* on macOS arm64. If you are the first
> to run it on Windows or Linux and no voice comes out, that's a bug worth reporting —
> not you doing it wrong.

## Build from source

Needs Node and Rust; for offline, [Ollama](https://ollama.com) running.

```bash
npm install
npm run tauri dev      # develop
npm run tauri build    # package
```

## More

- **[What's inside](./docs/architecture.md)** — the design, the speech ladder, the
  trade-offs we took and what they cost.
- **[Contributing](./CONTRIBUTING.md)** — especially language packs. Adding a language
  is a folder and one import line.
- **License:** [MIT](./LICENSE). Fork it, bundle it, ship it inside your own tool.

verba

File Explorer

# Use via CDN

Was this content helpful?

Recent Feedback

Download README