marker vs docling
Document conversion tools compared
marker and docling convert PDFs and other documents into structured content for downstream use. marker emphasizes configurable conversion with local model-based layout and OCR processing, while docling covers a broad range of document and media formats and offers integrations with AI and retrieval frameworks.
marker: Convert Documents into Structured Text
Marker converts PDFs and other documents into Markdown, JSON, HTML, or chunks, preserving structure such as tables, equations, and images. It suits developers building document-processing workflows who can run its local models and inference backend.
docling: Convert Documents into Structured Content
Docling converts PDFs and many other document formats into structured representations and exports such as Markdown and JSON. It is suited to developers building document ingestion workflows for search, analytics, and generative AI, including local processing of sensitive files.
| marker | docling | |
|---|---|---|
| Language | Python | Python |
| License | Apache-2.0 | MIT |
| Stars | 40.2k | 68.3k |
| Forks | 2.9k | 5k |
| Last analyzed | Oct 3, 2026 | Oct 3, 2026 |
Key differences
- marker exports Markdown, JSON, HTML, or chunks; docling also lists WebVTT and lossless JSON among its export options.
- marker uses the Surya vision-language model through a local inference server, with setup and hardware requirements that depend on mode and workload; docling supports local and air-gapped execution, with capabilities dependent on the format and pipeline.
- docling lists audio parsing and integrations with LangChain, LlamaIndex, Haystack, and Crew AI; marker lists custom processors, renderers, providers, and optional LLM-based correction.
- marker is Apache-2.0 licensed, while docling is MIT licensed; both note that model licenses may differ from the project code license.
- marker has 40.2k stars and 2.9k forks, while docling has 68.3k stars and 5k forks.
- marker describes its included API server as suitable for small-scale use; docling offers API-server and MCP options.
Choose marker if you…
- need chunked output or configurable processors and renderers for a conversion workflow.
- want a local conversion pipeline using marker's balanced or fast modes.
- are prepared to review the separate model-weight license and configure local inference.
Choose docling if you…
- need parsing across document, image, and audio formats, or exports including WebVTT.
- want integrations with LangChain, LlamaIndex, Haystack, or Crew AI.
- need local or air-gapped processing and can select a pipeline suited to your input formats.
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.