In internal use today · public release Q2 2027
Independent cloud CAT · TM · QA · AI

Any file. Any AI.
One platform.

TM, terminology, QA and your whole team in one editor. Every file goes back in its own format. Your AI keys or ours — no seat fees.

Join the waiting list → See what it does
Built by linguists, on real client work — not demos.
  • €0 per seat, always
  • BYOK or metered tokens
  • 40+ file formats, in and out
  • Independent — no tool lock-in
  • EU company, Estonia
demo.docx EN → FR Medical · Consent live
Source
I understand that this procedure carries risks, including bleeding and infection.
Target
Je comprends que cette intervention comporte des risques, notamment des saignements et une infection.
Tab accept AI suggestion
Translation memory
96%Je comprends que cette intervention comporte des risques, notamment des saignements et une infection.
88%Je reconnais que cette intervention comporte des risques, dont des saignements et des infections.
74%Le patient comprend que toute intervention chirurgicale comporte des risques de saignement et d'infection.
✓
ResolvedMQM clean
Linguist applied the fix — the segment is now MQM-clean.
or watch it work
scroll ↓
02The vision

An independent translation platform, built for the AI era.

One cloud-hosted, audit-ready platform for terminology, AI translation and MQM review — linguists in charge at every step, billed by usage, not by seat.

Available now

In internal use

Live, not a demo. Our in-house team and the freelance and agency linguists we work with use locAGI daily, on real client projects.

The full CAT environment: editor, TM, terminology, comments, MQM review. Native — not bolted on.

Studio
Source/target editor TM concordance Terminology MQM review Comments & @mentions Live presence Autosave
AI
OpenAI Anthropic Google xAI Grok BYOK or our tokens AI Assist per segment All languages, one workspace
Enterprise
Owner · Admin · Manager Translator · Reviewer Audit log per project Activity trail Workspace cost visibility
Independent by design: not an add-on to another CAT tool. Bilingual files from any of them — XLIFF, MQXLIFF, SDLXLIFF, MXLIFF — open here and go back in their own format.
Public release · Q2 2027

Open sign-up

Free to sign up. Your whole team joins free; you pay only for the AI you use.

Scheduled for Q2 2027. Until then, internal use — join the waiting list ›

Free
Platform access 1 GB storage No card to start
AI
Pay-as-you-go tokens BYOK (own API keys) Direct token rates
Team
Unlimited seats All roles included No per-head billing
Roadmap

What we're building toward

Still in flight:

  • SSO (SAML + OIDC) · SCIM provisioning
  • SOC 2 Type II readiness · data-residency controls
  • Fragment assembly · domain-weighted TM fuzzy scoring
  • Configurable review workflows · vendor-portal UX
What's already real: integrated CAT editor, TM and terminology, MQM scoring, BYOK or metered AI, multilingual projects, audit logs, the alignment review tool, a REST API with per-user keys.
03Platform

One workspace from source file to signed-off delivery.

TM, terminology, AI post-editing and MQM review in one editor. Every change attributable.

Govern

Terminology and memory that bind the output

A project termbase and a running glossary bind every AI call: a term decided in batch 2 still holds in batch 40.

  • TBX, TMX, XLSX and CSV import
  • Per-client templates bundle rules, tone and termbase
  • Consistency enforced across long documents
Automate

AI post-editing on the model you choose

Your keys or our metered tokens. Fuzzy bands set how hard the model works, so you never pay to retranslate a match.

  • Any model: OpenAI, Anthropic, Google, Mistral, xAI, DeepL — or your own
  • Private models via Ollama or any OpenAI-compatible endpoint
  • Deterministic prompts, logged per batch
Assure

MQM review with an auditable trail

MQM Core, LISA QA 3.1 or your own categories. Findings carry severity, comment and fix, and travel back into the file.

  • Single-language and multilingual review grids
  • Findings export to Excel, per file or merged
  • Comment threads with @mentions and activity history
04Interoperability

Any file in. The same file out.

Client documents and other tools’ bilingual files in; each one back in its own format. No file prep, no lock-in.

  1. Source file
  2. AI post-edit
  3. Glossary
  4. Translated file
40+ file formatsDOCX, XLSX, PPTX, IDML, HTML, XML, DITA, JSON, PO, RESX, SRT and more, rebuilt in the original format
Bilingual files from any CAT toolMQXLIFF, SDLXLIFF and MXLIFF open as they are and go back in their own format
XLIFF 1.2 & 2.0Standards-compliant, multi-file aware
Project packagesmemoQ handoff packages come in whole and go back as delivery packages
Tags & locks keptInline tags, placeholders and locked segments survive the round trip
TMX & TBXBring existing memories and termbases with you
REST APIProjects, segments, QA runs and exports — see the docs
Excel findings exportSend the QA findings out as a spreadsheet
05Data & control

Your keys, your data, your jurisdiction.

Client confidentiality stays under your control, not ours.

Bring your own keys Your own provider accounts; their contract governs usage and retention.
Private model option A self-hosted or private OpenAI-compatible model: content never reaches a third-party AI.
EU company, EU jurisdiction Built and operated by Terratra OÜ in Tallinn, Estonia.
Export everything, always Files, memories, termbases and findings in open formats. Leaving is a download.
06Capabilities

Built for teams that deliver at scale.

The engine, the team workspace and the integrations behind every project.

AI engine

Consistent voice across long documents

Your termbase and a running glossary ride with every AI call, so terms and tone hold across 10,000 segments. Prompts are logged; failures retry per segment.

Termbase-aware Running glossary Logged prompts Per-segment retry
AI Assist

AI at your fingertips, segment by segment

One click and the AI rewrites the target with your glossary and style guide. Or open the segment chat and ask it why.

AI Improve Segment chat Predicted next Style-guide aware
Collaboration

Your whole linguistic team. Same project. Live.

In-house linguists, reviewers and freelancers, live in one project: presence shows who is on which segment, a team chat settles questions and terms. Split files, assign translation and review steps, share a secure vendor link.

Live multi-user editing Team chat & @mentions Decisions & polls File splits Vendor links
Multilingual review

One project. Every target language.

One source, every target, MQM review side by side: one shared TM, a scorecard per language, one rollup for the PM.

All targets in one project Shared TM Per-language scorecard One rollup
File formats

Drop in what your clients send

Word, Excel, PowerPoint, InDesign, HTML, XML, JSON and more, with no file prep — and other CAT tools’ bilingual files. Each goes back in its own format, tags intact.

.docx · .xlsx · .pptx .idml XLIFF 1.2 · 2.0 MQXLIFF · SDLXLIFF · MXLIFF TMX · TBX
Automation

REST API

Read projects and segments, save translations, pretranslate, run QA and export — wire locAGI into your TMS or CI.

REST & JSON API keys Webhooks OpenAPI spec
07Pricing

The platform is free. You only pay for AI.

€0 per seat, ever. Pay only for the AI you use — or bring your own keys. Pricing from the public release in Q2 2027.

Free forever
€0
platform · seats · setup · annual commit
  • Unlimited seats, freelancers included
  • No credit card to start
  • Storage included
Pay-as-you-go AI
Tokens
only what you consume · no monthly minimum
  • One transparent bill
  • No tiers to outgrow
  • Switch to BYOK any time
Bring your own keys
No cut
OpenAI · Anthropic · Google — on your card
  • Calls go straight to your vendor
  • No margin on your AI bill
  • Mix with pay-as-you-go
See the full breakdown — including what we don't pretend to have yet →

Free to start. Yours to control.

One independent platform for every file, every language and every linguist. Public release Q2 2027.

Join the waiting list See pricing