Skip to content

General FAQ

How is DevGlish different from Grammarly / Duolingo?

Section titled “How is DevGlish different from Grammarly / Duolingo?”

DevGlish is an English learning tool designed specifically for developers, built into your workflow — not a general language learning or grammar checking app.

AspectDuolingoDevGlish
Learning methodSystematic courses + exercisesContextual lookup + instant feedback
Use caseDedicated learning timeLook up anytime during work
Target usersAll language learnersDevelopers
DomainGeneral EnglishTechnical terms + developer expressions
Learning depthBroad grammar and vocabularyDeep technical context

Summary: Duolingo is an “active learning” tool; DevGlish is an “instant lookup” tool. The two complement each other.

AspectGrammarlyDevGlish
FunctionGrammar checking, style improvementWord lookup, pronunciation learning
UsagePassive checking (after writing)Active lookup (when encountering new words)
Pronunciation guidanceNoneYes (with IPA and audio)
Non-native pattern detectionNoneYes (Chinglish and other L1 interference detection)
Integration depthIntegrated in editors/emailGlobal hotkey (any app)

Summary: Grammarly corrects what you’ve already written; DevGlish helps you choose the right words before you write. Both can be used together.

DevGlish supports macOS 13.0 and later (including Ventura, Sonoma, Sequoia).

  • Apple Silicon (M1, M2, M3, etc.) ✓
  • Intel (x86-64) ✓

If your Mac runs macOS 12 Monterey or earlier, DevGlish is not supported. We recommend updating macOS, or contact us to discuss other options.

DevGlish uses the macOS Accessibility API to read text you select in any app.

When you select text in Xcode, Slack, Chrome, or any app and press ⌘⇧D, DevGlish uses the Accessibility API to capture the selected text, then:

  1. Sends it to the Claude AI for lookup
  2. Returns a learning card
  • DevGlish only reads text you actively select; it does not monitor in the background
  • Your text is processed at Cloudflare edge nodes with per-user isolation
  • Text is not used for model training (see below)
  • Permission can be revoked at any time (System Settings > Privacy & Security > Accessibility)

If you don’t trust this permission, you can choose not to grant it, but DevGlish will be unable to function.

Can I use it without an internet connection?

Section titled “Can I use it without an internet connection?”

Partially, yes.

  • 85+ pre-loaded tech terms: Stored locally, no network needed
  • Cached queries: Previously looked-up words are cached (Free: 7 days, Pro: 30 days)
  • New lookups: First-time lookup of a new word/phrase/sentence requires network (calls the Claude API)
  • Express Mode: Native language to English conversion requires network
  • Advanced TTS: Multi-accent and speed control require network
  • Cloud sync: Saving words and syncing learning progress require network

If you plan to use DevGlish offline for more than 3 days:

  1. While online, look up words you might need in advance
  2. The system automatically caches these lookups
  3. While offline, you can access cached words and the local term database
  4. After more than 3 days offline, DevGlish automatically falls back to Free mode feature limits

Currently macOS only.

DevGlish is built with Swift and AppKit, technologies specific to the macOS and iOS ecosystem.

The following are on our roadmap:

  • VS Code Extension (estimated 2026 Q3): Look up directly in the code editor
  • Chrome Extension (estimated 2026 Q4): Quick lookup in the browser
  • Windows version (under evaluation): Possibly launching in 2027

If you need cross-platform support, please vote on the feedback form.

Is my data safe? Are query texts used for AI training?

Section titled “Is my data safe? Are query texts used for AI training?”

Your data is safe, and query texts are not used for model training.

  1. During lookup: Your selected text is sent to Cloudflare Workers edge nodes
  2. Proxy processing: The DevGlish backend acts as a proxy for the Claude API, forwarding your query
  3. Model call: Claude returns structured learning card data
  4. Result storage: Learning cards are stored in your Cloudflare D1 database (per-user isolation)
  • Text is not stored in Claude’s logs for model training
  • Text is not used to fine-tune or improve Claude
  • Cloudflare does not use your data for any other purpose
  • Your saved words and learning progress are accessible only by you (end-to-end encryption option available in Pro)

We use Anthropic’s official Claude API and follow their Privacy Policy. Per the policy, Anthropic does not use API inputs to train models unless you explicitly opt in. DevGlish never opts in.

For most developers, yes.

Based on our user research:

  • The average developer looks up 15-30 words per day
  • High-frequency users (writing lots of English PRs or documentation) look up 30-50 words per day
  • Even “heavy English users” rarely exceed 100 lookups/day

You may want to upgrade if:

  • You look up more than 10 words per day
  • You need Express Mode (native language to English)
  • You need to sync learning progress across multiple Macs
  • You need the full spaced repetition system for long-term review

Otherwise, the Free plan is usually sufficient.

Are DevGlish and DevLingo the same product?

Section titled “Are DevGlish and DevLingo the same product?”

Yes, they are the same product.

  • DevGlish is the product’s brand name and official English name
  • DevLingo is the original development name and internal technical name
  • Early 2025: Project started under the codename “DevLingo”
  • Mid 2025: Rebranded to “DevGlish” (more intuitively conveys the product’s function)
  • Now: The official name is “DevGlish,” but code, documentation, and internal discussions may still use “DevLingo”

In this documentation, both names may be used interchangeably, but they refer to the same product.

We’d love to hear from you!