Skip to content
MakerGain
← Back to project

Napkin Replaces the Canvas With a Paste Box, Finding Visuals in Text You’ve Already Written

Edited by MakerGain

Napkin makes the entry point to visual generation a paste box: users paste in text they’ve already written, the system offers candidate visuals, and users then adjust color, font, and layout. The team moved from Osmo’s children’s AI games to a text-first approach because the “draw first” barrier was too high. The free tier offers 500 credits per week; Plus costs $9 per person per month. Slides first asks about the goal, audience, and takeaway, then generates an editable presentation, but mobile still cannot generate visuals.

Napkin’s entry point is not a blank canvas but a paste box. Paste in a piece of text you’ve already written, click generate, and the system offers several candidate visuals based on the text; you pick one, then change colors, fonts, layout, and icons, and individual elements can be pulled apart and reshaped. The website explicitly states: no prompt writing required. [1]

This entry point came from an attempt that didn’t work out. The team founded Osmo about ten years ago, making AI games for children’s learning; in its About us retrospective, they first tried using AI to recreate two people doodling together on digital paper. After multiple rounds of iteration, they realized that even if AI drew better, “draw it first” was still a high barrier for most of their target users. Once LLMs became stronger, a text-first entry point became feasible, and the team shifted to generating visuals from text, which became Napkin’s starting point. [2]

Image: Source Napkin AI actual image source[3], accessed 2026-11-10; original interface retained.

For example, you’ve just finished writing a new process and need to put it into weekly meeting materials, but you’re missing a diagram that lets people understand the relationships at a glance. Napkin Visuals only presents candidate visuals; the color scheme, fonts, and layout are still yours to decide. It doesn’t write the content for you, and the choices remain in your hands.

The team’s own retrospective explains better why they chose this entry point: they were tired of a large number of 10-page documents and 100-page slide decks, and they believed visuals could make communication more effective, but most people don’t know where to start, or how to make it persuasive, so they can only write text. About us - Napkin AI[2] In a 2024 podcast, Pramod Sharma put the problem more directly: in presentations, blogs, or LinkedIn posts, graphics are always a large part, but making the right graphic takes a lot of time. Pramod Sharma - Founder of Napkin AI Talks Simplifying Visuals for Presentations[4] Together, these two statements define Napkin’s product positioning: it doesn’t teach people to draw, and it doesn’t require them to learn design first; it finds visuals in text that already exists.

The product scope also grows around this positioning. The website lists four use cases: presentations, blogs, social media, and documents; file import supports DOC, PDF, PPT, Markdown, and HTML, with unlimited file imports on the free tier. After generation, you can export PPT, PNG, PDF, and SVG; the free tier has unlimited PNG/PDF export, while PPT/SVG export starts with Plus. [1] · Pricing - Napkin AI[5] To take a visual into PowerPoint, Google Slides, Canva, or Keynote and continue editing fonts, colors, sizes, and animations, the exported PPT is what connects to these tools. [6]

Where the Money Is Charged

Pricing separates the free tier from the paid tier: Free is $0, with 500 AI credits per week; you can generate and edit visuals, and import files and export PNG/PDF without limits, but visuals carry Napkin branding. Plus is $9 per person per month, with 10,000 AI credits per month, and adds PPT/SVG export, removal of branding, and 3 brand styles. Credits are roughly counted at 1 point per word in the text you select for generation. The free allowance runs by week: when this week’s 500 is used up, you can wait for the next week for it to reset; to keep generating immediately, you can upgrade or buy extra credits as described on the pricing page. [5]

The company has not disclosed revenue, the number of paying customers, or its cost structure, nor does it provide data on generation quality or success rates for the same input. What can be confirmed is the plans on the pricing page, and the company’s statement in the Slides announcement: over the past two years, millions of users have used Napkin, generating more than 100 million visuals cumulatively. These numbers describe usage, not paid conversion; profit also remains unclear. [7]

First Ask Three Questions, Then Give You a Presentation

Slides is the next step in the same logic. In that announcement, the company said users often put visuals made in Napkin into presentations, and then ask Napkin to make the entire presentation directly. Slides does not start with a canvas; it first asks three things: what the goal is, who it’s for, and which point you want the audience to remember. The answers determine the outline, structure, content, and visuals; each page can generate multiple versions that are detailed or brief, text-driven or image-driven, all editable. Napkin AI - Napkin Slides: The Next Chapter[7] This process puts “think it through first” before generation, consistent with Napkin’s logic of starting from text. Slides is currently in beta, included in existing subscriptions at no extra charge. [7]

The company attributes the timing to three things getting stronger at once: LLMs got stronger at understanding context and reasoning, agents can handle multi-step workflows, and image models have also improved significantly. This is the company’s explanation, with no further technical details. [7]

There is also a gap the website itself acknowledges: on mobile, you can view existing work and create and edit presentations, but visual generation is unavailable; the website says desktop is still the best experience for generating visuals, and no timeline has been disclosed. Napkin AI - Turn Text into AI Diagrams and Visuals[1] For a product that treats generation as its entry point, this is not a peripheral feature; it means you can edit a presentation on your phone, but you cannot grow a visual from a new piece of text.

What Napkin is really betting on is not how good a single visual looks, but whether people who write materials are willing to hand “where to start” over to a paste box; Slides then pushes the question back to users: what is the goal, who is it for, and which point do you want the audience to remember. Whether these two entry points can connect in the same workflow says more about how far it can go than whether a single visual looks good.

Continue reading

AI productivity
An Excel file in, a software list ready for sign-off 90 days later

Before renewal, a public-sector organization was unsure whether its software was in use. Samplify’s consolidation task starts with the customer’s Excel file of publishers and products and three workshops. After 90 days, it delivers a sourced product-by-product list—keep, replace, maybe, defer—with three rounds of customer sign-off on key confirmations. Founder James Layfield says the company consolidates by feature and team and helps employees transition, but the Sprint delivers only analysis and a roadmap; migration and contract exits are a separate delivery.

AI productivity
He gave AI conversations a tree for the moments they need a detour

Slashspace handles the urge to pursue a side question while keeping the main thread intact when reading a long document: any answer can branch into its own node while the main conversation stays in place. Founder Praneeth Pike moved the product from a web subscription to a one-time desktop purchase, using BYOK to put model-call costs on users’ own keys and server costs on their computers.

AI images
A hundred thousand signups at zero marketing cost: Supermeme finds the template first

Supermeme reverses meme-making: users describe a situation first, then the system finds templates by meaning and drafts editable captions. It takes over template browsing and the first caption; people still judge tone and potential offense. In that interview, Sanjeev said the three founders had never met offline, collaborating on WhatsApp and twice-monthly Zoom calls, with largely reactive product discovery and instinct-led building decisions. Strictly speaking, they had yet to go all in and each had other projects. Sanjeev remained a product manager until November 2022, then explored Supermeme full-time while consulting to pay the bills.

AI productivity
An assistant cost thousands of US dollars a month. He put AI to work and reached \$10k+ MRR

Because an assistant would cost thousands of US dollars a month, Elie Steinbock built an AI email assistant instead. It connects to an inbox, sorts messages, and drafts replies using email history, calendars, and customer records; a person still presses Send. On Product Hunt launch day, he emailed 20,000 newsletter subscribers and posted a video on his YouTube channel with 2,000 subscribers. He estimated that these extra efforts may have added fewer than 50 votes—a judgment, not a measured result.