Skip to main content

Kapi Desktop

Kapi Desktop is the visual companion for the kapi CLI. It provides a native desktop interface for building localization flows, running tools, managing plugins, and storing AI credentials — all without writing YAML or remembering CLI flags.

Like the CLI, the desktop is project-first: it opens to your projects and reopens the ones you had last time, so a .kapi project — its languages, content, flows, translation memory, and termbase — is the home you work from. One-off file runs still live a click away as quick tools (the ad-hoc mode), for when you just want to push a single file through a tool without setting up a project.

A quick tour​

Create and manage a project​

A project pairs a source language with the locales you translate into, the files in scope, and the flows that process them. Naming it is the only required step — everything else is configured visually. The project home shows each content collection's translation status — how many blocks are covered per locale — and lets you extract content into the project store to refresh it.

Manage content in a project​

A project's content is the set of files it translates. Glob patterns are grouped into named collections — documentation, store UI, contracts, templates — each mapping source files to their translated counterparts. The Content view shows the patterns alongside the real files they match and the format Kapi detected for each.

Compose flows​

A project's flows are the pipelines it runs over its content — a plain AI translation, a translation followed by quality checks, a pseudo-translation for testing. Each flow is a named sequence of steps, saved alongside the project and shared through version control. Open one to see and edit its pipeline.

Configure appearance, providers, and plugins​

Settings gather everything in one place: light or dark appearance, the AI providers Kapi translates with (keys stay in your OS keychain), and the plugins that extend its formats and tools.

Browse terminology and translation memory​

Kapi keeps a glossary of multi-locale terms and a memory of past translations. Both are searchable, with inline formatting and entity placeholders preserved.

Using the Okapi plugin​

Plugins extend Kapi with new formats and tools. The Okapi bridge adds the Okapi Framework's filters — dozens of document formats beyond the built-in set. Install it from the registry in one click, and projects built on those filters, like the OkapiMart sample, open and process their files through them.

Features​

  • Project dashboard — Open to your projects and pick up where you left off. Each project home shows its languages, content collections, and per-locale translation status, with flows a click from running.
  • Flow editor — Build multi-tool pipelines visually. Chain AI translation, quality checks, pseudo-translation, and more into reusable flows saved in the project.
  • Flow runner — Execute a project's flows with live progress visualization: per-file status, node highlighting, and timing. Runs use the recipe's languages and content, so there are no flags to repeat.
  • Quick tools (ad-hoc) — For one-offs, run an individual tool or flow on a loose file with dynamic configuration forms generated from tool schemas — no project required.
  • Plugin manager — Browse the plugin registry, install, update, and manage plugins from a UI.
  • Credential vault — Store AI provider API keys securely in your OS keychain.
  • Project files — The .kapi recipe is a portable file you can share via git or open with double-click.

Install​

On macOS, the cask installs the app and bundles the kapi CLI:

brew install --cask neokapi/tap/kapi

For the Windows installer and the macOS DMG / Linux tarball builds, see Installation.

Your first project​

The tour above shows each surface in motion; here is the path through them when you start from nothing. Everything is configured visually — there is no YAML to write by hand.

  1. Create the project. Launch Kapi Desktop and click New Project. Give it a name, set the source language (e.g. en-US), and add the target languages you translate into (e.g. fr-FR, de-DE). Naming it is the only required step.
  2. Add content. In the Content view, add a pattern that points at your source files — for example src/locales/en/*.json. Kapi lists the files it matches and the format it detected for each, grouped into a named collection.
  3. Add a provider key. In Settings → Providers (or Credentials), add an AI provider and paste your API key. Keys are stored in your OS keychain, never in the project file.
  4. Compose a flow. In Flows, create a flow and add the steps you need — an AI translation, optionally followed by a quality or terminology check. Each flow is a named pipeline saved alongside the project.
  5. Run it. Select the flow, pick input files and a target language, and click Run. You'll see per-file progress, node highlighting, timing, and the output file locations.
  6. Install plugins (optional). In Plugins, browse the registry and install the Okapi Bridge to add its document formats and filters; new formats appear automatically.

Everything you configure is written to a portable <name>.kapi recipe file. Commit it to git to share with your team, reopen it in Kapi Desktop, or run it from the command line:

kapi run translate -p myproject.kapi

Next Steps​