Gamma, the AI presentation startup, lets users generate polished decks directly inside ChatGPT, part of a wider strategy to embed its design tools in the AI assistants people already use. Rather than copying an outline into a separate platform, a user can describe what they need in plain language, and Gamma turns it into a structured presentation.
Although the pitch has drawn fresh attention this week, the ChatGPT app is not brand new: Gamma’s own changelog shows it launched on March 6, alongside a similar connector for Anthropic’s Claude and native integrations with the automation tools Zapier, Make and n8n.
The integration runs on Gamma’s connector built on the Model Context Protocol, or MCP, an open standard that lets AI assistants call outside tools. In practice, ChatGPT handles the thinking, such as summarizing a report or structuring an argument, while Gamma handles the visual design and returns a finished artifact.
After a one-time setup through the ChatGPT app directory, a user can ask it to turn a technical report into an executive summary, build a marketing brief, or create a short pitch deck. Output is not limited to slides; the connector can also produce documents, webpages and social posts, with control over theme, tone, text density and export format. One clear limit is that the integration only creates content. Editing must happen back in Gamma’s own editor, so the workflow is to generate a first draft in the chat, then refine it in the app.
Gamma’s ambition is to become a visual output layer that plugs into many assistants rather than a standalone destination. Its integration list now spans ChatGPT, Claude, Atlassian Rovo, Superhuman and the enterprise search tool Glean, each aimed at a place where work already accumulates, from emails and meeting notes to project documents and support threads. The company’s framing is that a presentation usually starts from existing material, so the tool should live where that material is created.
Why It Matters
The connector targets a real friction point: turning finished thinking into a presentable format has long meant copying text into a deck builder and formatting it by hand. Removing that step is genuinely useful for the marketers, founders, consultants and sales teams who produce decks constantly.
The approach also reflects a broader shift in AI tools, from generating text toward completing whole workflows. Instead of only writing slide copy, an assistant can now produce the finished deck. And it shows how general assistants and specialist design tools are choosing to interoperate through MCP rather than compete head-on, each doing the part it does best.
The Competitive Question
The strategy carries an obvious risk: the assistants Gamma plugs into are building their own presentation features. OpenAI, Anthropic and others are steadily adding native document and slide generation, and Anthropic has its own design-focused tools, which could erode the need for a third-party layer over time.
Gamma’s bet is that its design quality, templates and editing environment are strong enough that users will still route generation through it, even when a built-in option exists. Being available across many platforms via an open standard is a hedge against dependence on any single one. Whether that positioning holds as the big model makers expand their own output capabilities is the central question for Gamma and the many startups pursuing a similar connector-first approach.
