MCP Servers · PDFMathTranslate

PDFMathTranslate

PDFMathTranslate is a Python tool that translates PDF scientific papers while preserving formatting, formulas, and layout. It supports multiple translation backends (Google, DeepL, OpenAI, Ollama) and offers CLI, GUI, Docker, and plugin interfaces.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/PDFMathTranslate/PDFMathTranslate

35.5k

GitHub stars

3.2k

Forks

Python

Primary language

AGPL-3.0

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	PDFMathTranslate/PDFMathTranslate
Owner	PDFMathTranslate
Primary language	Python
License	AGPL-3.0 — OSI-approved
Stars	35.5k
Forks	3.2k
Open issues	139
Latest release	v1.9.11 (2025-07-11)
Last updated	2026-05-25
Source	https://github.com/PDFMathTranslate/PDFMathTranslate

What PDFMathTranslate is

Built on ONNX-based document layout detection (DocLayout-YOLO), the tool parses PDFs into structured elements, translates content via pluggable service backends, and reconstructs PDFs with original formatting intact. Supports Python 3.11–3.12 with pip/uv installation and containerized deployment.

Quickstart

Get the PDFMathTranslate source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/PDFMathTranslate/PDFMathTranslate.git2cd PDFMathTranslate3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Multilingual Research Paper Access

Academic teams needing rapid translation of foreign-language scientific papers while preserving mathematical equations, citations, and table structure for citation and reference workflows.

Cross-Lingual Document Management Integration

Organizations using Zotero or other document management systems that need automated PDF translation as part of literature review pipelines, with preserved formatting for archival and sharing.

Enterprise Translation Workflow

Technical organizations requiring on-premise or API-driven PDF translation with support for multiple translation backends, Docker containerization, and custom service integration (Ollama, OpenAI).

Implementation considerations

Python version constraint (3.11–3.12 only) limits deployment flexibility; verify target environment compatibility before adoption.
Dependency on external model (`wybxc/DocLayout-YOLO-DocStructBench-onnx`) from HuggingFace; network issues may require mirror configuration (HF_ENDPOINT environment variable).
Translation quality and cost vary significantly by backend selection (Google vs. DeepL vs. OpenAI); test with representative PDFs early to validate accuracy for your domain.
AGPL-3.0 license mandates source disclosure if offered as a service; internal-only deployments have fewer restrictions but cloud/SaaS use requires licensing review.
GUI mode runs on localhost:7860 via Gradio; production deployments should reverse-proxy with authentication and rate-limiting.

When to avoid it — and what to weigh

Proprietary/Closed Commercial Software Requirement — AGPL-3.0 licensing requires source code disclosure if deployed as a service; commercial use must either remain internal or apply for alternative licensing arrangements with maintainers.
Unsupported PDF Formats or Scanned Images — Tool is optimized for text-based PDFs with embedded fonts and layout metadata; scanned or image-heavy documents require OCR preprocessing not built into this tool.
Real-Time Low-Latency Translation — Model loading (DocLayout-YOLO) and translation service latency make this unsuitable for sub-second translation requirements; batch processing is the intended workflow.
Sensitive/Regulated Data Without Careful Backend Selection — Default integration with cloud services (OpenAI, Google, DeepL) requires careful review of data residency and compliance policies; on-premise Ollama option exists but requires infrastructure overhead.

License & commercial use

Licensed under AGPL-3.0 (GNU Affero General Public License v3.0). This is a copyleft license requiring any modifications or derivative works, including network service deployments, to make source code available to users. Internal use and modification for private purposes are permitted; distribution or service provisioning triggers disclosure obligations.

AGPL-3.0 permits internal commercial use without restriction, but prohibits commercial service offerings (SaaS, hosted API) without either: (1) open-sourcing all modifications and derivative infrastructure, or (2) negotiating a separate commercial license with the copyright holder. Contact maintainer (Byaidu) for commercial licensing terms if you intend to offer this as a hosted or embedded service.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Adequate
License clarity	Clear
Deployment complexity	Moderate
DEV.co fit	Good
Assessment confidence	High

Security considerations

AGPL-3.0 requires source visibility if deployed as a service, supporting auditability but complicating proprietary deployments. PDF processing involves loading user-supplied documents and external models; validate file sources. Translation backends (OpenAI, Google, DeepL) transfer content over HTTPS; review backend data handling policies for sensitive research. Ollama option enables fully offline operation for regulated environments. No security audit or CVE history visible in provided data.

Alternatives to consider

Paperless-ngx / Ocrmypdf + Translation Layer

Document management focused; requires custom translation integration but offers broader workflow support (OCR, annotation, archival). Better for organizations already invested in document systems.

LLM-based PDF Rewriting (Claude API, GPT-4 Vision)

Native PDF vision support avoids layout reconstruction complexity; better for image-heavy or complex layouts. Higher per-page cost but simpler integration. Lacks scientific formula preservation guarantees.

Paid SaaS Solutions (e.g., Adobe Acrobat AI, Transifex)

Commercial support, compliance guarantees, and simplified deployment. Trade-off: vendor lock-in, ongoing subscription costs, and less technical control over translation backend selection.

Software development agency

Build on PDFMathTranslate with DEV.co software developers

Translate scientific papers instantly while preserving every equation, chart, and citation. Try the free online demo or deploy locally with Docker.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

pdf-reader-mcp markdownify-mcp spacy-layout

Related on DEV.co

Explore the category and the services that help you build with it.

MCP Servers AI Application Development API Development & Integration Cloud Deployment & DevOps Talk to DEV.co

PDFMathTranslate FAQ

Can I use this in a commercial SaaS product?

Not without modification or licensing. AGPL-3.0 requires open-sourcing your service infrastructure. Contact the maintainer for commercial licensing, or use Ollama as an on-premise backend if deploying internally only.

What translation quality should I expect?

Quality depends on your chosen backend (DeepL typically highest, Google mid-range, Ollama depends on local model). Test with representative papers in your domain before production rollout. Scientific terminology may require post-processing.

Does it work offline?

Yes, using Ollama as the translation backend. However, the DocLayout-YOLO model must be downloaded first (internet required). After that, full offline operation is possible if Ollama is locally hosted.

What Python versions are supported?

Python 3.11 and 3.12 only. Python 3.10 and earlier, or 3.13+, are not supported. Verify your environment before installation.

Software development & web development with DEV.co

Need help beyond evaluating PDFMathTranslate? DEV.co is a software development agency offering software development services and web development for teams of every size. Our software developers and web developers build custom software, web applications, APIs, and mcp servers integrations — and maintain them long-term.

Streamline Your Research Workflow

Translate scientific papers instantly while preserving every equation, chart, and citation. Try the free online demo or deploy locally with Docker.

Talk to DEV.co Schedule a Consultation