GhostEdit

Fix your writing anywhere. No cloud required.

Cross-platform AI text correction from the menu bar

Why GhostEdit?

	GhostEdit	Grammarly	LanguageTool	Apple Writing Tools
Works in any app	Yes (system-wide)	Browser + select apps	Browser + select apps	Yes
Offline mode	Yes (built-in Bonsai model)	No	Server only	Yes
Multi-provider AI	Local, Claude, GPT, Gemini, any OpenAI-compatible API	Grammarly AI only	LanguageTool only	Apple AI only
Open source	MIT	No	LGPL	No
Price	Free	$12/mo	$5/mo (premium)	Free (Apple only)

Key Features

Works in every app -- System-wide hotkey correction, not a browser extension
Offline-first AI -- Built-in Bonsai grammar model runs on-device, no API keys needed
Multi-provider -- Switch between local AI, Claude, GPT, Gemini, Grok, or any OpenAI-compatible API in one click
Bring your own key -- OpenAI, OpenRouter, Groq, Together, Grok (xAI), Claude/GPT/Gemini native endpoints, Ollama, LM Studio, and Custom presets, each with its own encrypted key (~/.ghostedit/api-key-<profile>.enc)
Real-time monitoring -- Traffic light indicator shows writing quality as you type
Developer-friendly -- Preserves code, URLs, @mentions, :emoji:, and file paths
13 languages + tone presets -- Auto-detect language, choose from 5 writing styles

Quick Start

Download the latest release from GitHub Releases
Open GhostEdit -- it lives in your menu bar
Select text anywhere and press the hotkey: Cmd+E (General provider) / Cmd+Shift+E (local) on macOS, Ctrl+... on Win/Linux

Features

Global hotkey — Two configurable shortcuts: Cmd+Shift+E / Ctrl+Shift+E (local) and Cmd+E / Ctrl+E (General-tab provider)
Offline-first — Built-in Bonsai model served by the bundled llama.cpp runtime runs entirely on-device, no API keys needed
Cloud providers — Claude, Codex (OpenAI), and Gemini via their CLI tools
OpenAI-compatible API — Bring your own key per provider: OpenAI, OpenRouter, Groq, Together AI, Grok (xAI), Claude/Codex/Gemini native endpoints, or local servers (Ollama, LM Studio); multi-profile setup with per-provider curated models, Refresh discovery, and encrypted key storage
Dictionary pre-pass — Harper.js (grammar) + nspell (spelling) fix obvious errors before the AI sees the text, making corrections faster and cheaper
Dictionary polish — Same engine runs again on model output to catch any remaining issues
Diff preview — Side-by-side streaming diff before accepting changes (Tab to accept, Esc to cancel)
Token preservation — URLs, @mentions, :emoji:, `code`, email addresses, and file paths are never modified
Tone presets — Default, Casual, Professional, Academic, and Slack styles
13 languages — Auto-detect or specify: English, Spanish, French, German, Italian, Portuguese, Japanese, Korean, Chinese, Russian, Arabic, Hindi
Correction history — Browse, search, and review past corrections
Correction cache — Identical inputs skip the AI entirely
Bundled inference runtime — Pinned llama.cpp server per OS/architecture, SHA-256 verified at build time
3 Bonsai model sizes — 1.7B (248 MB, default), 4B (572 MB), and 8B (1.1 GB), downloaded on demand
System tray — Runs silently in the menu bar with no dock icon (macOS)

Requirements

Node.js >= 18
npm >= 9
Operating System: macOS 12+, Windows 10+, or Linux (x64/arm64)
Accessibility permissions (macOS): Required for keyboard simulation — System Settings > Privacy & Security > Accessibility > enable GhostEdit

Optional (for cloud providers)

Provider	CLI Tool	Install
Claude	`claude`	claude.ai/cli
Codex	`codex`	`npm install -g @openai/codex`
Gemini	`gemini`	ai.google.dev/gemini-api/docs/cli

No CLI? Add an API key per provider for any OpenAI-compatible host (OpenAI, OpenRouter, Groq, Together AI, Grok/xAI) or a local server (Ollama, LM Studio) in Settings > Providers — see AI Providers. Keys are stored encrypted, one file per provider (~/.ghostedit/api-key-<profile>.enc, e.g. api-key-grok.enc).

Cloud providers are not required. The llama.cpp server is bundled per OS and architecture. When Local is selected, GhostEdit downloads the Bonsai model from Hugging Face to ~/.ghostedit/models/bonsai/; after download, local corrections work offline.

Installation

Clone and install dependencies

git clone https://github.com/nareshnavinash/GhostEdit.git
cd GhostEdit
npm install

Each installer bundles the matching llama.cpp CPU server executable and its native libraries. When the user selects Local during onboarding, GhostEdit downloads the selected Bonsai GGUF model from Hugging Face into ~/.ghostedit/models/bonsai/. API/CLI users do not download the local model unless they choose to download it in Settings > Local Model.

Quick Start

# Start the app in development mode
npm start

On first launch, onboarding lets you choose a provider before local model setup:

Choose a provider: Local (offline), an OpenAI-compatible API preset (each preset keeps its own model, base URL, and encrypted API key — set up as many as you need, including keyless Ollama/LM Studio), or a cloud CLI tool.
For Local, pick a Bonsai model size (1.7B / 4B / 8B) — the app downloads it and shows progress.
For an API preset, pick the preset card, enter your base URL, model (curated dropdown + Refresh from host), and API key (stored encrypted per profile).
Set preferences and try a correction in the onboarding flow.

After setup:

Select text in any application
Press the hotkey
A HUD overlay shows "Working..."
The corrected text replaces your selection (or appears in a diff preview)

Usage

Basic correction

Select text in any application (editor, browser, Slack, email, etc.)
Press the hotkey (Cmd+Shift+E local / Cmd+E General-tab provider by default)
GhostEdit captures the selection, corrects it, and pastes the result back

Diff preview mode

When Show diff preview is enabled (default):

A preview window opens showing original vs. corrected text side-by-side
Insertions are highlighted in green, deletions in red
Press Tab to accept, Escape to cancel, R to regenerate

Clipboard-only mode

When enabled, corrected text is copied to clipboard instead of being pasted back. Useful when paste simulation doesn't work in a specific app.

System tray menu

Right-click (or click on macOS) the tray icon to access:

Correct (Local) — Run the offline model (Cmd+Shift+E)
Correct (API) — Run the General-tab provider (Cmd+E)
Undo Last Correction — Revert the last paste (Cmd+Shift+Z)
Settings... — Open the configuration window
History... — Browse past corrections
Quit GhostEdit — Exit the app

Configuration

All settings are stored in ~/.ghostedit/config.json and editable through the Settings window.

Settings tabs

General — Language, tone, and correction preferences

Local Model — Bonsai model size, download, and inference device

Providers — Per-CLI model + path cards, plus per-API-profile cards (model dropdown with curated + Refresh-discovered models, base URL, encrypted API key per provider)

Hotkeys — Record custom global shortcuts (local, General provider, undo, line)

Behavior — Toggle features:

Setting	Description	Default
Fast correction mode	Greedy decoding for local model (faster, slight quality trade-off)	On
Clipboard-only mode	Copy to clipboard instead of pasting back	Off
Show diff preview	Side-by-side preview before applying	On
Sound feedback	Play sound on completion	On
Notify on success	System notification on correction	Off
Developer mode	Show inference device info in tray and settings	Off
History limit	Max stored corrections	50

Tone presets

Preset	Style
Default	Standard grammar/spelling correction
Casual	Friendly, conversational, keeps contractions
Professional	Polished business communication
Academic	Formal vocabulary, precise structure
Slack	Concise, upbeat, preserves emoji and abbreviations

Custom system prompt

Create ~/.ghostedit/prompt.txt to override the default system prompt. The file is read on each correction.

Config file locations

File	Purpose
`~/.ghostedit/config.json`	Application settings
`~/.ghostedit/history.json`	Correction history
`~/.ghostedit/prompt.txt`	Custom system prompt (optional)
`~/.ghostedit/api-key.enc`	Legacy single encrypted API key (auto-migrated; removed when cleared)
`~/.ghostedit/api-key-<profile>.enc`	Per-provider encrypted API key (e.g. `api-key-openai.enc`, `api-key-grok.enc`)
`~/.ghostedit/models/bonsai/`	Downloaded Bonsai GGUF models
`~/.ghostedit/device-cache.json`	Cached GPU/CPU detection result

AI Providers

Local model (default)

The default engine is Bonsai, a grammar-correction GGUF model served by the llama.cpp CPU server bundled with every installer (pinned release, SHA-256 verified at build time for each OS/architecture).

Model sizes (downloaded on demand from Hugging Face to ~/.ghostedit/models/bonsai/, then fully offline):

Model	Size	Notes
Bonsai 1.7B	248 MB	Default — best speed/quality balance
Bonsai 4B	572 MB	Better quality
Bonsai 8B	1.1 GB	Best quality

Pick a size in Settings > Local Model or during onboarding. The first Local correction downloads the selected model; afterwards corrections work with no network access.

A legacy T5 pipeline (transformers.js + ONNX) remains as a fallback for old configs and is auto-migrated to Bonsai on first launch. When it is active, GPU acceleration is automatic:

Platform	Primary Device	Fallback
macOS (Apple Silicon)	WebGPU (Metal)	WASM (CPU)
Windows (any GPU)	DirectML	WebGPU / WASM
Linux x64 (NVIDIA)	CUDA	WebGPU / WASM
Linux arm64	WebGPU	WASM (CPU)

OpenAI-compatible API (bring your own key — one key per provider)

GhostEdit v1.9.0 gives every API preset its own profile: model, base URL, and encrypted API key are saved independently, so you can configure OpenAI, Grok, Ollama, and others all at once instead of one global key.

Pick presets in Settings > Providers — each card shows a Configured / Not configured badge. Hosted providers become configured once their key is saved; keyless hosts (Ollama, LM Studio, Custom) become configured once you pick/confirm a model. Paste your API key (encrypted at rest with Electron safeStorage into ~/.ghostedit/api-key-<profile>.enc), choose from the curated model dropdown or press Refresh to list models live from the host.

Preset	Base URL	Default model	Key required?	Curated models
OpenAI	`https://api.openai.com/v1`	`gpt-4.1-mini`	Yes	`gpt-4.1-mini`, `gpt-4.1`, `gpt-4.1-nano`, `gpt-4o`, `gpt-4o-mini`, `o4-mini`, `o3`
OpenRouter	`https://openrouter.ai/api/v1`	`openai/gpt-4.1-mini`	Yes	`openai/gpt-4.1-mini`, `openai/gpt-4.1`, `anthropic/claude-sonnet-4`, `anthropic/claude-3.5-haiku`, `google/gemini-2.5-flash`, `meta-llama/llama-3.3-70b-instruct`
Groq	`https://api.groq.com/openai/v1`	`llama-3.3-70b-versatile`	Yes	`llama-3.3-70b-versatile`, `llama-3.1-8b-instant`, `gemma2-9b-it`, `mixtral-8x7b-32768`
Together AI	`https://api.together.xyz/v1`	`meta-llama/Llama-3.3-70B-Instruct-Turbo`	Yes	`meta-llama/Llama-3.3-70B-Instruct-Turbo`, `meta-llama/Llama-3.1-8B-Instruct-Turbo`, `Qwen/Qwen2.5-72B-Instruct-Turbo`
Ollama	`http://localhost:11434/v1`	`llama3.2`	No	`llama3.2`, `llama3.1`, `llama3.1:8b`, `mistral`, `gemma2`, `qwen2.5`
LM Studio	`http://localhost:1234/v1`	`local-model`	No	`local-model`
Custom	(your URL)	(your model)	No	(your saved model + Refresh results)
Claude (Anthropic API)	`https://api.anthropic.com/v1`	`claude-sonnet-4-5`	Yes	`claude-sonnet-4-5`, `claude-opus-4-1`, `claude-haiku-4-5`, `claude-3-7-sonnet-latest`, `claude-3-5-haiku-latest`
Codex (OpenAI API)	`https://api.openai.com/v1`	`gpt-5-codex`	Yes	curated Codex/GPT list
Gemini (Google API)	`https://generativelanguage.googleapis.com/v1beta`	`gemini-2.5-flash`	Yes	curated Gemini list
Grok (xAI API)	`https://api.x.ai/v1`	`grok-3-mini`	Yes	`grok-3-mini`, `grok-3`, `grok-2-1212`

Then pick which configured provider actually runs in Settings > General > Default correction model — GhostEdit only offers providers you have configured (local is always offered; CLIs only when installed; APIs only when configured). The API Hotkey (Cmd/Ctrl+E) runs whatever is selected there.

Corrections stream into the diff preview like every other provider.
API keys are encrypted at rest (~/.ghostedit/api-key-<profile>.enc, legacy api-key.enc auto-migrated) and never stored in config.json.
Legacy single-key configs migrate automatically: your old key/base-URL/model move into the profile you had selected.
Trigger with Ctrl/Cmd+E (General provider) or the tray's Correct (API) entry.
Local servers (Ollama, LM Studio) and Custom bridges work without an API key. Custom endpoints even fall back to the SSE streaming path for browser-backed bridges.

Cloud providers

Cloud providers use their respective CLI tools, spawned as subprocesses. You must install and authenticate the CLI separately.

Claude

# Install
# Visit https://claude.ai/cli

# Authenticate
claude auth login

# GhostEdit auto-detects the CLI, or set the path in Settings

Codex (OpenAI)

npm install -g @openai/codex
codex auth

Gemini

# Install from https://ai.google.dev/gemini-api/docs/cli
gemini auth

Dictionary Engine

GhostEdit includes a two-layer dictionary engine that mirrors the approach from the companion macOS-native app:

How it works

Harper.js (primary) — Rust-powered grammar checker compiled to WASM. Catches spelling, grammar, capitalization, and style issues with suggestions.
nspell (secondary) — JavaScript port of Hunspell with a full English dictionary. Fills non-overlapping gaps that Harper misses.
Merge strategy — All Harper issues are kept. nspell issues are added only if they don't overlap with any Harper issue. Proper nouns (capitalized mid-sentence) and acronyms (2+ uppercase letters) are filtered out.
Iterative passes — The engine runs up to 3 passes. Fixing one error can reveal another (e.g., fixing punctuation may expose a grammar error). Stops early when no more changes are found.

Where it runs in the pipeline

Selected text
  -> Token protection (URLs, @mentions, code, emoji)
  -> Dictionary pre-pass (fix obvious errors)     <-- HERE
  -> Cache lookup
  -> AI engine (local Bonsai / OpenAI-compatible API / cloud CLI)
  -> Token restoration
  -> Dictionary polish (cleanup model output)      <-- AND HERE
  -> Paste back

The pre-pass means most simple typos ("teh" -> "the") never reach the AI model, making corrections faster and reducing API costs for cloud providers.

Architecture

+-----------------------------------------------------------+
|                   Electron Main Process                     |
|                                                             |
|  Global Hotkey --> Clipboard Capture --> Token Protection    |
|       |                                                     |
|       v                                                     |
|  Dictionary Pre-pass (Harper + nspell, up to 3 passes)      |
|       |                                                     |
|       v                                                     |
|  Cache Lookup --[hit]--> Skip AI                            |
|       |                                                     |
|       v [miss]                                              |
|  Correction Dispatcher                                      |
|    +-- CLI Runner (claude/codex/gemini subprocess)          |
|    +-- API Runner (OpenAI-compatible + native APIs, stream) |
|    +-- Local Runner (llama.cpp + Bonsai GGUF server)        |
|    +-- Inference Window (WebGPU/WASM T5 fallback)           |
|       |                                                     |
|       v                                                     |
|  Token Restoration --> Dictionary Polish --> Paste Back      |
|       |                                                     |
|       v                                                     |
|  History Store --> Correction Cache                          |
|                                                             |
|  System Tray | IPC Handlers | Device Selector               |
+-----------------------------------------------------------+
                          | IPC
+-----------------------------------------------------------+
|                  Renderer Process (React)                    |
|                                                             |
|  Settings | History | HUD Overlay | Streaming Preview        |
|           | Onboarding Wizard | Inference Worker             |
+-----------------------------------------------------------+

Key design decisions

Tray-only app — No dock icon on macOS (LSUIElement: true). Runs silently in the menu bar.
Context isolation — Renderer processes are sandboxed. All main/renderer communication goes through a secure preload bridge.
Graceful degradation — If Harper fails to load (WASM issue), nspell still works. If both fail, the AI model handles everything. If the inference window crashes, the main process pipeline takes over.
Fire-and-forget pre-warming — Model, dictionary checkers, and keyboard simulator are loaded in parallel at startup without blocking the UI.
Correction cache — Keyed on (text, provider, model, tone, language). Invalidated on any config change.

Development

Start in development mode

npm start

This launches Electron with Vite HMR — changes to renderer code are hot-reloaded instantly. Main process changes require a restart.

Type checking

npx tsc --noEmit

Linting

npm run lint

Environment

The project uses:

TypeScript 5.7 with strict mode
Vite 6 for building main, preload, and renderer
React 19 for the UI
Tailwind CSS 3 with a custom dark theme
Zustand 5 for renderer state management
Vitest 4 for testing

Testing

# Run all tests
npm test

# Watch mode
npm run test:watch

# Run a specific test file
npx vitest run src/main/dictionary-checker.test.ts

# Run with verbose output
npx vitest run --reporter=verbose

Test structure

File	Type	Tests	Covers
`dictionary-checker.test.ts`	Unit	89	Harper/nspell extraction, merge, filtering, fix application, initialization
`dictionary-checker.integration.test.ts`	Integration	36	End-to-end with real Harper WASM and nspell Hunspell
`token-preservation.test.ts`	Unit	60	URL, mention, code, and emoji protection
`constants.test.ts`	Unit	35	Provider/preset definitions and defaults
`Settings.test.tsx`	Component	50	Settings sections, model downloads, per-profile API key management
`local-model-runner.test.ts`	Unit	24	Pipeline loading, device routing, streaming, variant switching
`config-manager.test.ts`	Unit	27	Config persistence, T5 → Bonsai migration, hotkey + per-profile API migration
`history-store.test.ts`	Unit	18	History persistence and limits
`openai-compatible-runner.test.ts`	Unit	12	API presets, model discovery, streaming, native endpoints
`ipc-handlers.test.ts`	Unit	14	IPC handler registration and behavior (incl. per-profile keys)
`tray-manager.test.ts`	Unit	20	Tray menu entries (Local + API, no dedicated CLI item)
`device-selector.test.ts`	Unit	11	Platform detection, disk caching, DirectML/CUDA probing
`correction-dispatcher.test.ts`	Unit	12	Provider routing (local / CLI / API + per-profile config)
`global-shortcuts.test.ts`	Unit	12	Hotkey registration and conflict detection (2 correction hotkeys)
`cli-runner.test.ts`	Unit	5	CLI subprocess spawning and error handling
`provider-shortcut.test.ts`	Unit	5	General-tab provider → shortcut target resolution

Current: 530 tests across 28 files, all passing.

Building & Packaging

Package (unsigned)

npm run package

Creates an unpacked app in out/.

Build installers

npm run make

Produces platform-specific installers:

Platform	Format	Output
macOS	ZIP on Linux; ZIP + DMG on macOS	`out/make/`
Windows	Squirrel (`Setup.exe`) on Windows; ZIP on other hosts	`out/make/`
Linux	Debian	`out/make/deb/x64/`

Build a Debian package

On Linux x64, install dpkg-deb and fakeroot, then run:

npm run make:deb

The packaging hook downloads the pinned llama.cpp CPU archive for the target, verifies its SHA-256, and includes its server executable and libraries in the .deb. The Bonsai model remains an on-demand Hugging Face download. Debian dependencies for Electron and keyboard/desktop integration are declared in the package metadata.

Linux ARM64 builds can be made with:

npm run make:deb:arm64

Build Windows and macOS installers

Build on Windows for the native Squirrel installer. Linux and macOS builds produce a ZIP of the Windows application instead; it contains the app executable but not the Squirrel install/update workflow.

# Windows x64: Squirrel installer on Windows, ZIP on Linux/macOS
npm run make:win

# Windows ARM64: Squirrel installer on Windows, ZIP on Linux/macOS
npm run make:win:arm64

# macOS Apple Silicon: ZIP on Linux, ZIP + DMG on macOS
npm run make:mac:arm64

# macOS Intel: ZIP on Linux, ZIP + DMG on macOS
npm run make:mac:x64

Windows produces the Squirrel installer set, including Setup.exe, when built on Windows. Cross-builds from Linux or macOS produce a ZIP instead. For macOS targets, Linux builds produce only a ZIP; macOS builds produce both ZIP and DMG. Each build bundles only its own OS/architecture server.

What gets bundled

The Electron app (ASAR-packed)
The target OS/architecture llama.cpp server and libraries
App code and UI assets (Bonsai model weights are downloaded per-user)
All production dependencies
App icon (assets/icon.svg)

Project Structure

src/
  main/                          # Electron main process
    index.ts                     # App lifecycle, correction pipeline, window management
    config-manager.ts            # Config persistence (~/.ghostedit/)
    correction-dispatcher.ts     # Routes corrections to local, CLI, or API providers
    cli-runner.ts                # Spawns CLI subprocesses
    openai-compatible-runner.ts  # OpenAI-compatible + native API corrections (streaming)
    api-key-store.ts             # Encrypted API key storage (Electron safeStorage)
    local-model-runner.ts        # Legacy T5 pipeline (fallback)
    llama-runtime-manager.ts     # Bundled llama.cpp server resolution
    llama-server-manager.ts      # llama.cpp server lifecycle
    bonsai-model-manager.ts      # Bonsai GGUF download and scanning
    dictionary-checker.ts        # Harper.js + nspell spell/grammar checking
    clipboard-manager.ts         # Cmd+C/V simulation via nut.js
    token-preservation.ts        # Protects URLs, @mentions, code, emoji
    correction-cache.ts          # In-memory correction cache
    history-store.ts             # Persists correction history to JSON
    tray-manager.ts              # System tray icon and context menu
    inference-window.ts          # Hidden BrowserWindow for GPU inference
    device-selector.ts           # GPU/CPU detection and caching
    global-shortcuts.ts          # Global hotkey registration
    ipc-handlers.ts              # IPC event handlers
    cli-arguments.ts             # CLI path resolution
    error-messages.ts            # User-friendly error messages

  renderer/                      # React UI
    App.tsx                      # Root component (routes by window type)
    main.tsx                     # React entry point
    windows/
      Settings.tsx               # Configuration UI
      History.tsx                # Correction history browser
      HudOverlay.tsx             # Transparent status overlay
      StreamingPreview.tsx       # Real-time diff preview
    components/
      Welcome.tsx                # Onboarding wizard
      DiffView.tsx               # Side-by-side diff viewer
      HotkeyInput.tsx            # Hotkey recording input
      ProviderSelector.tsx       # Provider selection buttons
      ModelSelector.tsx          # Model dropdown
    inference-worker.ts          # Web worker for model inference

  preload/
    index.ts                     # Secure IPC bridge (contextBridge)

  shared/
    types.ts                     # TypeScript interfaces and IPC channels
    constants.ts                 # Providers, models, languages, defaults

assets/
  icon.svg                       # App icon

resources/
  bin/                           # llama.cpp server runtime per platform-arch (downloaded at build)
  models/                        # Legacy T5 model download target (optional)

scripts/
  download-llama-server.mjs      # Pinned llama.cpp runtime download (SHA-256 verified)
  download-bonsai-model.mjs      # Bonsai GGUF download helper
  download-model.mjs             # Legacy T5 model download script

Troubleshooting

"No text was captured"

Make sure text is selected before pressing the hotkey
On macOS, grant Accessibility permission: System Settings > Privacy & Security > Accessibility > enable GhostEdit
Try clipboard-only mode if paste simulation fails in a specific app

"CLI: Not found"

Install the CLI tool for your chosen provider (see AI Providers)
Or set the CLI path manually in Settings > General

"Authentication required"

Run the auth command for your provider:
- Claude: claude auth login
- Codex: codex auth
- Gemini: gemini auth

Local model is slow

Pick the smaller Bonsai 1.7B model in Settings > Local Model
Enable "Fast correction mode" in Behavior tab
Use a cloud CLI or the OpenAI-compatible API provider for heavier edits
Check the inference device in Settings with Developer Mode enabled

Corrections take too long

Increase the timeout in Settings > General
Switch to a faster model (e.g., haiku for Claude, o4-mini for Codex)
Use the local model for instant offline corrections

Hotkey doesn't work

Check for conflicts with other apps using the same shortcut
Change the hotkey in Settings > Hotkey tab
On macOS, ensure the app has Accessibility permissions

Model download fails

Check your internet connection and retry from Settings > Local Model
Models are downloaded from Hugging Face — ensure huggingface.co is accessible
Files land in ~/.ghostedit/models/bonsai/; delete a partial download and retry

"API request failed"

Verify the right profile's API key in Settings > Providers (each provider stores its own encrypted key in ~/.ghostedit/api-key-<profile>.enc)
Confirm the base URL and model name match your host — use Refresh on that provider's card to reload the list
For Ollama / LM Studio, ensure the local server is running on the expected port
The HUD and error log show the HTTP status and message returned by the host
If your Custom bridge reports conversation_deleted, reset its saved conversation and retry

What's New in v1.9.0 — Per-Provider API Keys + Simplified Hotkeys

Your API-key system, explained: every provider preset (OpenAI, OpenRouter, Groq, Together, Grok/xAI, Claude/Codex/Gemini native, Ollama, LM Studio, Custom) now keeps its own model + base URL + encrypted key. Configure several providers at once — the Providers tab shows a Configured / Not configured badge per card. A new Grok (xAI) preset (https://api.x.ai/v1, grok-3-mini + curated list) joins the matrix, each preset ships a curated model dropdown plus Refresh discovery, and legacy single-key configs auto-migrate into the profile you had selected.

Hotkeys are simpler: 2 correction shortcuts instead of 3 — Cmd/Ctrl+Shift+E (Local) and Cmd/Ctrl+E (General-tab provider, resolved via the new provider-shortcut module). The tray drops the dedicated CLI item in favor of Correct (Local) + Correct (API). Pasting is safer via replaceSelectedText() (re-verify → Backspace → paste), each CLI remembers its own model, the proofreading system prompt no longer answers embedded questions, and the onboarding installs the local model before the Try-it step.

File (v1.9.0)	Platform	Size
`GhostEdit-darwin-arm64-1.9.0.zip`	macOS Apple Silicon	~480 MB
`GhostEdit-darwin-x64-1.9.0.zip`	macOS Intel	~483 MB
`GhostEdit-win32-x64-1.9.0.zip`	Windows x64	~492 MB
`GhostEdit-win32-arm64-1.9.0.zip`	Windows ARM64	~490 MB
`ghostedit_1.9.0_amd64.deb`	Linux x64 (Debian)	~322 MB
`ghostedit_1.9.0_arm64.deb`	Linux ARM64 (Debian)	~318 MB

Each bundle includes the pinned llama.cpp runtime; Bonsai models download on demand to ~/.ghostedit/models/bonsai/.

License

See LICENSE for details.

Name		Name	Last commit message	Last commit date
HomebrewFormula		HomebrewFormula
assets		assets
docs		docs
scripts		scripts
src		src
.gitignore		.gitignore
README.md		README.md
forge.config.ts		forge.config.ts
index.html		index.html
index.js		index.js
package-lock.json		package-lock.json
package.json		package.json
postcss.config.js		postcss.config.js
tailwind.config.ts		tailwind.config.ts
tsconfig.json		tsconfig.json
vite.main.config.ts		vite.main.config.ts
vite.preload.config.ts		vite.preload.config.ts
vite.renderer.config.ts		vite.renderer.config.ts
vitest.config.ts		vitest.config.ts

Latest commit

History

Folders and files

Repository files navigation

GhostEdit

Why GhostEdit?

Key Features

Quick Start

Table of Contents

Features

Requirements

Optional (for cloud providers)

Installation

Clone and install dependencies

Quick Start

Usage

Basic correction

Diff preview mode

Clipboard-only mode

System tray menu

Configuration

Settings tabs

Tone presets

Custom system prompt

Config file locations

AI Providers

Local model (default)

OpenAI-compatible API (bring your own key — one key per provider)

Cloud providers

Dictionary Engine

How it works

Where it runs in the pipeline

Architecture

Key design decisions

Development

Start in development mode

Type checking

Linting

Environment

Testing

Test structure

Building & Packaging

Package (unsigned)

Build installers

Build a Debian package

Build Windows and macOS installers

What gets bundled

Project Structure

Troubleshooting

"No text was captured"

"CLI: Not found"

"Authentication required"

Local model is slow

Corrections take too long

Hotkey doesn't work

Model download fails

"API request failed"

What's New in v1.9.0 — Per-Provider API Keys + Simplified Hotkeys

License

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages