Patentext May 2025 update: new editor experience, improved outputs, and more

We just wrapped up our first in-person event at the AIPLA Spring Meeting in Minneapolis (conveniently, where our co-founder Marcus lives). This was our first booth, our first public demos, and, frankly, a lot of firsts all around.

Over the course of three days, we ran dozens of demos on real invention disclosures, handed out plenty of stickers and flyers, and had some truly energizing conversations with patent professionals across the industry. 

Since getting back, we've been heads-down translating that momentum into product work, and I'm excited to share that we’ve just shipped our 0.2.0 release. It's our biggest update yet, and it makes Patentext faster, clearer, and easier to use across the board.

What’s new in Patentext 0.2.0?

This release focused on quality-of-life improvements — the kind that make drafting less frustrating and more fluid. If you’ve demoed Patentext in the past, you’ll notice the difference immediately.

A revamped editor experience

We rebuilt the editor from the ground up to make it faster, smoother, and easier to navigate:

• Easier block management: Add, delete, and reorder paragraphs more smoothly.
• Visible paragraph numbers: Makes reviewing and referencing much easier — and provides context for your position in the document.
• Draft outside of topic blocks: Add boiler plate, draft disclaimers, or define terminology without committing text to a section.
• Move content between topics: Shift things around as your draft evolves without having to copy and paste manually. 

There are also a lot of subtle UI tweaks throughout the app (e.g., layout spacing, visual cues) that make the whole experience feel more polished.

Smarter generation across the board

We’ve made major improvements to Patentext’s drafting engine, resulting in cleaner language, tighter structure, faster generation, and significantly less cleanup:

• Stronger detailed descriptions: Patentext now produces more complete paragraphs with better flow, improved references to figures, and fewer awkward transitions.
• Full application support: You can now generate draft content for every major section, including Claims, Abstract, Background, Summary, and Brief Description of Figures.
• More natural component descriptions: We’ve upgraded attribute-level generation to produce clearer, more intuitive language, even for complex or nested parts of the invention.
• Works even better with your invention data: Just build your graph, and Patentext will handle the structure, references, and phrasing without requiring prompt templates or manual reformatting.

New references feature

This was one of the most common requests we heard, so we built it. The new References feature uses named entity recognition (NER) to help you manage internal references more efficiently across the draft:

• Assign aliases to nodes: Standardize terminology across the application without having to hunt and replace manually.
• Auto-update figure numbers: Keep figure references consistent, even as your application evolves.
• Clickable internal references: Jump directly from a mention in the text to the associated node (rather than scrolling around trying to find “element 302.”)

What’s next?

Next up: more control over style, improved graph generation process, and tools that make it easier for teams to collaborate across clients and roles.

If you’ve been following along on our journey but haven’t scheduled a demo yet, early access is still open — and still discounted at $200/month. This includes full access to our tool, live support, and a direct line to the team as we keep building.

Schedule a demo today.

‍