FineTuner
Fine-tune AI models on your data — in minutes, not days.
179 followers
Fine-tune AI models on your data — in minutes, not days.
179 followers
FineTuner lets you train your own AI on your content (PDFs, websites, YouTube videos...) without writing a single line of code. Generate a high-quality dataset, fine-tune your model (GPT or Claude), and deploy it via API ready to use in minutes.
Free Options
Launch Team / Built With
AWS Activate CreditsCredits that fuel your startup’s next milestone
Promoted
👋 Hey Product Hunt!
I’ve always dreamed of an AI that could absorb my own content and speak in my exact voice. Two months ago I finally asked Cursor:
“Can you write a script that automates fine-tuning GPT-4o on my own content?”
That innocent prompt kicked off an 8-week rabbit-hole:
100+ workflows tested, 16-hour days, $250 in OpenAI bills, countless Docker tantrums.
I’m a low-code guy, so I had to learn real backend stuff on the fly—async Python, routing, deployments, the lot. Every time the AI “almost” worked, dopamine kept me glued to the screen.
Yesterday the last bug finally gave in. The result is Finetuner:
⚡ What it does ⚡
Point it at your Docs, PDF, YouTube, websites — anything.
In minutes you get a fully fine-tuned GPT-4o (or Claude) that speaks exactly in your voice.
No YAML rituals, no JSON wrangling. Just upload, click, done.
🛠 Why it’s different 🛠
Built 100 % with Cursor. Yes, the same AI that writes the code now teaches your model.
Handles the whole pipeline: data scraping ➜ cleaning ➜ training ➜ validation ➜ ready-to-use API.
Works for clones (“Tweet like Steve Jobs”) or personal brands (your unique tone).
Try it 👉 finetuner.io
I’d love feedback from the PH community: features you’d want, edge-cases you’re worried about, or horror stories from your own fine-tuning adventures. Drop a comment or DM me. Happy to share the gritty details of building a “real” app entirely with AI.
Thanks for hunting, and may the (Force) fine-tunes be with you!
@maxime_marsal1 Hi Maxime,
I’m Nima, co-founder of Bestofweb .site - a curated hub for remarkable digital tools and the makers behind them. Huge congrats on launching Finetuner, especially after pushing through that intense 8-week deep dive. Love how you’ve made fine-tuning feel effortless—no YAML or JSON nightmares, just upload and go. We’re supporting your launch and would be thrilled to feature your work in our directory.
Really like the idea and congratulations on the launch. Do you have any ideas on how much time it saves a company compared to training with code?
Hi Maxime,
First, congrats on the idea. It is quite good. Here is a little feedback which I hope will help you.
As there are already many similar tools, I would try to convey more efficiently what distinguishes you from the rest.
As you refine your project, I would also suggest you make your product compliant with data privacy, tax, etc., regulations (especially the EU ones). Those are real risks that most people coming up with the "next great" idea often don't consider.
In your FAQ 'data security', the wording you use implies that the data is not secure. I would actually consider spending a few hours learning about those things because your AI-generated FAQ is not really reassuring.
"The security and confidentiality of your data are our top priority. = Good
Your imported content is used solely for the purpose of creating your dataset and fine-tuning YOUR model, on YOUR provider account (OpenAI/Anthropic). = I read it as meaning that you don't secure it, but you don't train a model on it? That is nice.
We never resell your data or use your information to train other models. = Indeed, that would be illegal to do it without consent.
You retain full intellectual property rights over your content and the resulting fine-tuned model." = Same here, it would be illegal.
Hope this can help you improve your communication and ensure you cover your back from a legal standpoint.
Good luck!
Really cool concept! Love that FineTuner lets you fine-tune models without code. As someone in the creative field, I’m curious - can I use this to build a personalized assistant that helps me generate better prompts for AI art or storytelling projects? Would love to explore that use case!
Hey @maxime_marsal1, this is awesome and it's 100% better than using custom instructions/memories + prompting to get GPT to generate outputs in the right "brand voice" 👍
Quick question - once your model is fine tuned, this can be accessed via the dashboard you built and an API and an embedded chat that could be put on say, my personal website?
Thanks!
@ohansemmanuel Thanks a lot! I really apreciate!
FineTuner makes fine-tuning AI models accessible—great innovation! 😄