Emojis are showing in my requests, I specifically told Gamma not to do this. This is for the exportable pptx feature.

The overall API request payload looks like this:

create_payload = {

"inputText": full_prompt,

"textMode": "generate",

"format": "presentation",

"themeId": POWERPOINT_THEME,

"numCards": slide_amount,

"cardSplit": "auto",

"additionalInstructions": additional_instructions,

"exportAs": "pptx",

"textOptions": {

"amount": "medium",

"tone": "professional, analytical",

"audience": "You are speaking to business stakeholders.",

"language": "en",

},

"imageOptions": {

"source": "noImages",

},

}

full_prompt just has healthcare data in it that's not very important to the powerpoint generation.

And my additional instructions are as follows:

instructions = """Note that you are talking to pharmaceutical manufacturer stakeholders, not providers. Make the title informative, only utilize a title and nothing else should go at the top. Also no need to mention the steps in the title nor text blocks near the title, just never mention steps in general. Try not to have a lot of empty space in slides, make analysis font readable (not too small). Also utilize box formats for text and other formats when possible, so essentially try to avoid having the entirety of a slide just be text on a white background. The 4 slides are - "Educate Providers to Monitor" which explains what pharmacy companies should learn from patients pre-diagnosis signals, "Distribution Focus" which showcases what pharmacy should prioritize distribution to, "Broader Care Considerations" which should showcase a broader care paradigm, and "Account Strategy" which should indicate what medical centers to focus on and what referral pathways are optimal`. NEVER use emojis/emoticons/any icon images.'

The hosted Gamma MCP connector (used in Claude Desktop/Cowork) exposes only the generate tool, which starts generation AND polls for completion in a single call. When generating presentations with more than 3 slides, this takes over 30 seconds, which exceeds the MCP tool timeout in Claude environments.

The open-source Gamma MCP server on GitHub already solves this with a separate get_gamma_generation tool that checks status independently. Could the hosted connector add a similar tool?

Proposed solution - split into two tools:

This way each individual MCP call stays well under the timeout limit, and the AI assistant can poll get_generation_status until the deck is ready.

Currently: 3-slide presentations work, anything larger times out. We had to use the browser UI as a workaround. This would make the MCP connector reliable for presentations of any size.

Environment: Claude Cowork (Anthropic), Gamma Pro account, MCP tool timeout ~30 seconds.

Добавить возможность загружать свою тему не только через интерфейс приложения, но и через API. Это позволит использовать API Gamma в CRM системе, когда готовится презентации в шаблонах клиента.

I am facing issue related to publishing the site, for the gamma I created. There is no API endpoint for publishing the site.

The documentation currently explains to use 'noImages' when you wish to only add the URL's of your hosted images. - This is untrue, noImages literally takes out all the images from the slides. Please either create a URL's only tag 'customURL' or something similar and

regards,

AI developer and your future customer

Hi Gamma team,

I'm a Pro subscriber and active API user. I run an automation agency in France and I'm building automated content generation workflows for my clients using your API (via Make.com and n8n).

I'd like to request the addition of Studio Mode as a parameter in the Generate API (v1.0).

Currently, the API only supports the Classic card creation mode. However, Studio Mode where each card is a single AI-generated image with integrated text produces significantly more engaging content, especially for social media formats (LinkedIn carousels, Instagram posts, etc.).

A simple addition like a "cardMode": "studio" parameter in the API request body would unlock this for automation users and open up a whole new category of use cases for your API.

This is currently the main blocker preventing me from scaling my clients' social media content production through Gamma's API. I'd love to know if this is on your roadmap.

Thanks for building a great product.

Adam <3

Developer Documentation Specs - do it, we need it.

I get you're trying to do the whole "our agent can do anything" kinda route but at least set up some baseline levels of communication for your users to utilize the AI to the best of it's abilities.

I tried to do it manually for my GPT AI agent workflow, which took me 12 hours just talking and testing with your AI Helper Agent, and I'm afraid I mostly a wasted my time.

I have very specific brand guidelines and customization requirements for each slide, and without documentation my prompts are ineffective with your agent. I might as well just build it myself from GPT to Canva. At least it's more modular once I have a skeleton document.