Noplag Engine

Overview

Noplag Engine combines a local matching engine with hosted plans that add access to external corpora, including live-web search on Premium and Enterprise. Apache 2.0 licensing and Docker Compose deployment give technical teams room to run it on their own hardware; the trade-off is that they must manage the deployment themselves. Development requires Python 3.12 or newer, and the Compose setup uses PostgreSQL.

Key features

Winnowing fingerprints surface candidate sources, then a seed-and-extend aligner identifies matching character ranges. Reports break similarity down by source and include passage offsets and coverage accounting, so reviewers can locate matches rather than rely on a single score.

Corpus ingestion supports .txt, .md, .pdf and .docx files, including folder ingestion and HTTP uploads. The /v1 API covers checks, reports, uploads, progress events and corpus management, which can suit teams building checking into another workflow. It ships without authentication, however, so exposing it to users or a network requires careful security planning.

The engine does not detect paraphrase, semantic similarity or AI-written text. Those capabilities are described as roadmap work, not current functionality. The open-source engine searches a local corpus; the hosted product adds live-web search. Hosted detection is calibrated for English, Spanish, Portuguese, Polish and Ukrainian at launch.

The maker says submitted documents are not used to train models. Hosted-plan retention is configurable, and EU data residency is available on request—useful options for organizations assessing document handling.

Pricing

Free is 0.00 USD per free and allows three checks per month at up to 2,500 words per check, with public web and Common Crawl access. Self-hosting allows unlimited local checks, making it a practical route for evaluation or teams able to supply their own corpus and infrastructure; hosted use remains tightly capped.

Pro costs 9.99 USD per month, billed monthly, for 50 checks per month at up to 25,000 words each, plus the Noplag Database, academic index, REST API and webhooks. It fits makers and small teams that need recurring checks and integrations, but has no stated Live Web search.

Premium costs 19.99 USD per month, billed monthly, and raises the allowance to 200 checks per month and 50,000 words per check. Live Web search, branded PDF reports and priority support make it better suited to agencies and publishers with higher volumes or client-facing output.

Enterprise has custom pricing and custom check volume, with all corpora and Live Web, on-premises or private-cloud deployment, SSO/SAML, LMS integrations, dedicated support and an SLA. Its Canvas, Moodle and Blackboard integrations are joined by Brightspace via LTI 1.3. Those controls and deployment choices suit universities and regulated organizations, but require a custom engagement rather than a fixed-price subscription.

The pricing model also describes a 14-day free trial. The free plan’s three-check monthly allowance is distinct from that trial.

Platforms

Noplag Engine is offered for API, Linux, macOS, self-hosted, web and Windows environments. The repository’s stated development prerequisites—Python 3.12 or newer and PostgreSQL in Docker Compose—are relevant for teams maintaining their own instance.

Who it's for

Choose Noplag Engine when you need corpus-based verbatim matching, inspectable passage locations and an API, and have the technical capacity to manage self-hosting or secure an API deployment. It is less suitable when paraphrase or AI-text detection is central, or when a small monthly hosted quota cannot meet your checking volume.

Pros and cons

Pros

• Passage offsets, per-source similarity and coverage accounting make findings more actionable than a lone similarity score.
• Apache 2.0 licensing, Docker Compose deployment and unlimited local checks give technical teams control over where and how they run the engine.
• File ingestion and a broad /v1 API provide multiple ways to connect corpus checks to existing workflows.
• Configurable hosted retention and EU data residency on request address two document-handling concerns for organizations.

Cons

• No paraphrase, semantic-similarity or AI-writing detection, so it cannot cover those review needs.
• The API has no built-in authentication, placing security responsibility on the operator.
• Free hosted use is limited to three monthly checks of 2,500 words each; larger needs require a paid plan or self-hosting.
• Live Web search is reserved for Premium and Enterprise, limiting lower paid tiers to other listed corpora.

Alternatives

Noplag is a close alternative with the same free allowance of three checks per month and 2,500 words per check, plus public web and Common Crawl corpus access and Community Discord support; consider it if those support terms are a better fit.

Plagium offers five free daily searches with a 4,000-character limit and no Deep Search or API access, making it a different option for lighter searches.

GPTZero is the better direction when AI detection matters: its free plan includes AI detection, while Noplag Engine does not detect AI-written text.

JPlag is another free, open-source local option; choose it if Java SE 25 is suitable for your environment.

Originality.ai offers an AI-detection-only free plan with three scans per day and a 2,000-word scan limit, so it fits a narrower AI-detection need rather than corpus-based matching.

Pangram is an alternative for readers prioritizing AI detection, with a free allowance of 2,000 words per day, image detection scans, OCR, more than 20 languages, a browser extension and Google Docs integration.

Plagiarisma is another option.

Quetext AI Detector bundles an AI detector and plagiarism checker, each capped at 1,000 words on its free plan, alongside a summarizer and humanizer capped at 2,000 words.

For a broader comparison, see Plagiarism Detection Software.

Verdict

Noplag Engine is a strong fit for technically capable teams that want self-hosted or corpus-controlled verbatim matching, detailed passage reports and API access. Choose it for that control and traceability; look elsewhere if you need paraphrase or AI-text detection, or a larger hosted allowance without moving to a higher plan.