Presenton - Open Source AI Presentations - Locally generate presentations on your own design - UI & API
Presenton is an open-source (Apache 2.0) AI presentation generator that lets you create, edit, and export decks to PPTX and PDF.
Deploy instantly with Docker, then choose to run it locally with Ollama or connect to popular AI models using your own API keys.
Generate custom templates from existing slides or AI and use them through the user interface or API to create on design presentations, giving individuals, developers, and enterprises complete control over their AI presentation workflow.
Replies
Presenton - Open Source AI Presentations
Hi everyone,
We’ve been working on AI presentations for a while now. Over time, we realized how powerful they can be for communicating ideas clearly and efficiently. The need for AI presentation generators is huge, but current tools still miss a few key areas where this tech could really shine.
Many of the ideas people work on are sensitive. Especially for those in industries where privacy matters, having something that runs locally and not on someone else’s servers is really important.
We also noticed developers often want to generate presentations through an API, to integrate into their apps or automate customer assets. (Obviously with good documentation)
That’s why we built Presenton, and we’ll keep building on it for these reasons.
We’re not perfect, but we’re committed to adding value to the community and making something genuinely useful.
Would love to hear your thoughts and connect.
In the meantime, you can checkout docs for details about deployment and usage: https://docs.presenton.ai
Suraj
Swytchcode
Congrats on the launch! This is awesome. Love that you're taking the local-first and open-source route.
Curious, how do you handle the API side for generation? Is it a simple REST layer, or do you also support async/queue-based generation for larger decks?
Presenton - Open Source AI Presentations
@chilarai We do support queue based async generation.
You can checkout this https://docs.presenton.ai/guide/generate-presentations-asynchronously .
On our tests we've generated decks upto 50 slides with larger models like sonnet 4 and GPT 5.
This is super thoughtful @surajjha — especially the focus on privacy and control. Open-source, self-hostable with Docker, and the option to run locally with Ollama or plug in your own API keys feels spot-on for teams dealing with sensitive ideas.
Curious: how are people using the template + API combo so far — more for internal automation, or are you seeing it embedded into customer-facing tools as well?