GITMD2PDF is getting a fresh UI and more ways to fit into your workflow. Alongside the existing API and CLI, we’re adding an MCP server, and AgenticMods has built a Claude Code Mod. The next chapter will also be co-branded as RAG2PDF, reflecting the growing role of knowledge retrieval and AI workflows in creating documents.
Yes, AgenticMods has created a plugin Mod.
These changes bring GITMD2PDF closer to where documentation gets written: in repositories, in development tools, and increasingly in conversations with AI agents. Whether you start with a Markdown file, a repository full of documentation, or a Claude Code session, we want the journey to a useful, shareable document to feel straightforward.
GITMD2PDF started with a practical need. Markdown is a great format for writing and maintaining documentation, especially alongside code. But when that documentation needs to reach a client, a colleague, or someone reviewing a project, a PDF is often the easiest way to share it. It gives people a document they can open, read, and pass around without needing access to your development environment.
That remains the product’s foundation. The next release expands how you get there.
The revamped UI will give the web experience a fresh look and make the conversion workflow easier to navigate. For anyone who prefers to work in a browser, the goal is a clear path from source material to finished document. You should be able to bring in your Markdown, choose the options that matter, and get back to work.
The existing API and CLI continue to play an important role. They let GITMD2PDF fit into repeatable processes: generating documentation from scripts, adding PDF output to a publishing workflow, or creating documents from the terminal. Once you do a task regularly, a programmable way to run it makes a difference.
The API lets applications request conversions. The CLI brings that capability into the command line. Together, they already make GITMD2PDF useful beyond the browser and lay the foundation for new integrations.
The upcoming MCP server extends that access to AI tools. MCP, or Model Context Protocol, gives compatible assistants a standard way to discover and use external tools. For GITMD2PDF, that means an agent can include document conversion as part of a larger task.
Imagine asking an assistant to review a repository, prepare a project handover, and produce a PDF for the team. Or work through a technical proposal with an agent and ask it to turn the finished Markdown into a shareable document. The MCP server is being added to support those workflows, connecting content preparation with delivery.
This also creates an opportunity to reduce token use in your AI flow. An assistant creating a PDF through code may spend tokens writing a rendering script, adding formatting rules, reading errors, and revising the result. Once your Markdown is ready, GITMD2PDF can render it through a dedicated conversion service.
Let AI create the content. Let GITMD2PDF handle the PDF.
To put the potential savings into numbers, consider an illustrative workflow: an assistant generates 5,000 output tokens writing and revising PDF code, while delegating conversion takes 250 output tokens for a tool call and confirmation. That's 95% fewer output tokens for the conversion step.
At the published API output rates of $10 per million tokens for both Claude Sonnet 5.5 and OpenAI GPT-6.1 Sol, those assumed outputs cost $0.05 versus $0.0025. This is a worked example, not a measured benchmark. It excludes input tokens, content creation, tool execution charges, and GITMD2PDF plan costs; it is also not a per-PDF price for a Claude or ChatGPT subscription. Sources: Claude model pricing and OpenAI API pricing.
How you pass the content matters, too. Sending a long document through an MCP tool call still uses tokens. When supported, referencing an existing source can avoid reproducing its contents. The practical benefit comes from avoiding unnecessary rewriting and repeated PDF-generation code.
The RAG2PDF co-brand speaks to this broader direction. RAG, short for retrieval-augmented generation, uses relevant source material to inform an AI’s response. When that work produces a report, explanation, or technical document, it needs a practical path into a format people can read and share. RAG2PDF gives that use case a name alongside GITMD2PDF’s established focus on Markdown and repository documentation.
AgenticMods has also built a Claude Code Mod that brings document export directly into the Claude Code experience.
A productive coding session can contain much more than code. It can capture the reasoning behind a change, alternatives considered, implementation notes, and the context someone else will need later. That material is valuable when you’re handing work over, reviewing a decision, or returning to a project after some time away.
The Mod focuses on exporting that session material into Markdown, with PDF and HTML output through GITMD2PDF. By exporting the existing session text, it avoids asking Claude to regenerate the conversation just to create a document. GITMD2PDF then handles the rendering, reducing the need for additional model work during export.
This makes the Mod a useful companion throughout your AI workflow. Finish an investigation, export the findings, and share them with your team. Complete a coding session and preserve the discussion for a handover. Keep the knowledge you’ve already created without another round of document-generation prompts.
Each interface serves a different moment. The web UI suits a direct conversion. The API supports application integrations. The CLI fits terminal workflows. The MCP server opens access to compatible agents, and the Claude Code Mod brings session export into Claude Code.
We’re excited to bring these pieces together under the GITMD2PDF and RAG2PDF co-brand: a refreshed web experience, established developer tools, an upcoming MCP server, and a Claude Code Mod already built by AgenticMods. More ways to turn the knowledge and documentation you create into something other people can use, with less AI work spent getting it into PDF.