Logrus Global Launches a Free Tier of Perfectionist—A New Year Gift to Help the Industry

December 10, 2025 — As AI and machine translation become default tools in production, one uncomfortable truth is becoming clearer: fluent output is not the same as reliable output. In real workflows, quality can change with a prompt, a glossary update, a model version, or a routing decision. That variability creates operational, legal, and brand risk—especially when organizations publish at scale.

In manufacturing, the rule is simple: if you don’t measure, you don’t control. Translation Quality Evaluation (TQE) is that measurement layer for multilingual content. And when the content matters, TQE cannot be wishful thinking or “quick scoring”—it must be grounded in human judgment and executed as scientifically developed metric taking into account customer specifications and science of measurement, in a repeatable process.

That’s why Logrus Global built Perfectionist, a cloud-based Translation Quality Evaluation platform that helps teams run automatic and human-in-the-loop evaluations using a diverse portfolio of metrics—without the need to send Excel scorecards by email, or the overhead of building a bespoke QA program from scratch. Perfectionist combines an underlying database, process automation, and multiple metrics (Acceptance Testing / Acceptance Rate, LOGIPEM, ATA, GEMBA-SQM, and calibrated MQM scoring—linear and non-linear) so organizations can benchmark anything and measure quality frequently and consistently across languages, vendors, linguists, AI/MT engines, content types, use cases, and scenarios, or calibrate their other resources and methods.

Why the industry needs Perfectionist now

• Human MQM evaluation remains the gold standard. MQM provides a structured method to identify and quantify errors with severity and a scoring model—so results are stable, comparable, and actionable (not just a subjective “grade”). Perfectionist also implements decision-specific metrics: Acceptance Testing (Acceptance Rate) to select the right model per content type, LOGIPEM to distinguish between the quality of two AI/MT-generated datasets, the ATA metric to verify translator qualifications, and GEMBA-SQM for AI-only evaluation—alongside multiple MQM scoring options (linear calibrated and non-linear calibrated).
• AI workflows require more frequent measurement, not less. When outputs can shift, you need ongoing evaluation to detect drift, qualify changes, and keep risk under control.
• Statistically sound sampling makes human evaluation practical. You can get fast, reliable signal from a sufficient sample—often faster and more cost-effective than attempting to operationalize an “automatic metric” that still requires calibration to human judgment.
• You can benchmark anything—and choose the right model with evidence. Evaluate suppliers, compare MT vs LLM outputs, select the right model per content type via Acceptance Testing (Acceptance Rate), validate post-editing performance, track quality trends over time, and more.

Perfectionist also supports automated approaches where appropriate—but with one principle: automation must be calibrated against golden human judgment. Perfectionist is built to make that calibration measurable, repeatable, and scalable.

What’s new for 2026 before 2026 even started

After extensive development of new quality metrics, Logrus Global is introducing a free tier of Perfectionist. After self-registration, users can run up to three free evaluations per week, with no subscription required.

Users can apply standard MQM-based evaluation workflows, including linear-calibrated and non-linear-calibrated scoring. They can also use Acceptance Testing (Acceptance Rate) to select the right AI/MT engine per content type, LOGIPEM to compare outputs of different engines (especially raw vs post-edited), the ATA metric to verify translator qualifications, and GEMBA-SQM for AI-only evaluation where appropriate.

Availability
Perfectionist is available at https://perfectionist.app/
Media / partnerships: marcom@logrusglobal.com

About Logrus Global

Logrus Global is a language technology and localization company focused on enterprise translation quality, risk management, and measurable AI/MT adoption practices.