Connect Your Engine

Updated: 2 months ago · 2 min read

The Lingo.dev CLI connects your codebase to a localization engine through your project config. Every lingo push routes through your engine — applying your brand voice, glossary, rules, and model configuration automatically. No code changes, no new dependencies.

What changes#

BeforeAfter
Translation pipelineDefault Lingo.dev pipelineYour localization engine
Brand voiceNoneApplied per target locale
GlossaryNoneSemantically matched per request
RulesNoneIncluded per target locale
Model selectionLingo.dev defaultYour model config with fallbacks
Quality reviewNoneYour configured AI reviewers run automatically

Setup#

1

Link to your engine

Run lingo link in your project root:

bash
lingo link

Pick (or create) an organization and a localization engine. link writes orgId and engineId into .lingo/config.json — commit it so every run uses the same engine. Find an engine's ID in the dashboard too; it starts with eng_ (e.g. eng_SxjMwMsfOIsvV1wh).

2

Authenticate

bash
lingo login

Opens a browser to authenticate and stores a session. For CI or non-interactive use, pass an API key instead:

bash
lingo login --api-key lk_...

…or set it as an environment variable:

bash
export LINGO_API_KEY="lk_..."

Generate an API key from the API Keys page in the dashboard.

3

Push

bash
lingo push

The CLI sends source files to your engine and writes the translated results back to disk — applying your brand voice, glossary, rules, and model configuration automatically.

Configuration reference#

.lingo/config.json#

FieldDescription
orgIdOrganization that owns the engine. Set by lingo link.
engineIdEngine that performs the translation (eng_...). Set by lingo link.

sourceLocale, targetLocales, and files work the same as in any project — see Configuration.

Environment variables#

VariableRequiredDefaultDescription
LINGO_API_KEYFor CI-Your Lingo.dev API key (instead of lingo login)
LINGO_API_URLNohttps://api.lingo.devCustom API base URL (for self-hosted or staging)

What to expect#

Every translation the CLI produces flows through your engine's full pipeline:

  • Brand voice shapes tone and formality per locale
  • Glossary items enforce exact terminology via semantic matching
  • Rules apply locale-specific linguistic rules
  • Model configs select the right LLM with automatic fallbacks
  • AI Reviewers evaluate quality automatically after each request

Translations appear in Reports with trigger type api, alongside requests from the API and integrations.

Next Steps#