EnhanceAI - Add AI autocomplete to your website in 2 minutes | Product Hunt
source link: https://www.producthunt.com/posts/enhance-ai
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.
Add AI autocomplete to your website in 2 minutes
EnhanceAI
Add AI autocomplete to your website in 2 minutes
EnhanceAI autocomplete = more leads, reviews, complete profiles, and happy customers 💚
⚡️ Integrates with 2 lines of code (works on no-code too) 🛠 Customize your OpenAI model and prompt 🤑 Free up to 1M tokens
👉 enhanceai.dev 👈
👋 Hey! I’m Robert, one of the makers.
Thanks for checking out EnhanceAI - this is Walden’s and my first Product Hunt launch, so we’re a bit nervous 😄
Why we built EnhanceAI 🛠 Over the past few months, we and our friends built a TON of apps using GPT: email summaries, legal docs, you name it.
Every time we shipped something, we spent hours setting up and customizing AI boilerplate code 🤦♂️
Everyone should be able to easily add AI to their products, not just AI researchers and programmers!
What we do ⚡️ We make it ridiculously easy for anyone to add state-of-the-art language models to their product, starting with text inputs.
We integrate with all no-code apps and are free to use up to 1M tokens ✅
People use EnhanceAI to: 🤝 Reduce churn in their sales forms ⏱ Increase speed of using internal tools ✏️ Improve completion of long-form writing tasks 🧠 Help their users become more creative, professional, or funny 🔎 Analyze how their users use their AI products … and more!
We threw together a launch ASAP to get this to as many awesome people as possible. We’d love to meet you and hear your thoughts!
PS: If you beat Walden’s integration speedrun, we’ll double the tokens on your free plan! Fastest speedrun gets a surprise 👀
I've been chatting with Robert and Walden for a couple weeks while they've been honing in on this! So good to see this launched now and how easy it is to add ai autocomplete...
all my webflow sites can use this now !
Hey ProductHunt!
👋 I’m Walden, one of the makers and resident code monkey 👨💻
A Product Made with Love ❤️ The goal of EnhanceAI from the start was to make intelligence easily embeddable into any application
We pride ourselves in making a tool that is: - Ridiculously easy to use (for coders and non-coders alike) - Incredibly powerful out of the box
From when we went public with the first version, which was simply a React package, we already saw so many people using it for creative and surprising applications.
Seeing how many businesses and users loved EnhanceAI, we decided to make it even better, and available to the many people who use no-code websites.
The Technology 🤖 A little bit on the technology: we offer a script that takes any existing textarea and makes it AI-powered, suggesting text as you go, keeping everything else the same. This means: - No need to configure a bunch of CSS settings - Effortless for the user (no “+++” or “\ai”)
We put a lot of work into making it reliable across different browsers and platforms.
You can try us out in your application immediately. No account necessary! Let us know what you think and how we could continue to make the product better!
I've been dreaming about a drop-in product like this for years, ever since Gmail came out with their very good (until compared with GPT) autocomplete.
I asked for it a couple years ago: https://twitter.com/philfreo/sta...
EnhanceAI is the first thing I've seen that matches - really cool, congrats on the launch!
Which model is the homepage demo using?
How do you think about latency vs quality? My biggest concern for using OpenAI's APIs for this application was that autocomplete is most useful when you don't have to pause for a long time for it to predict.
@philfreo Thank you for the high praise! That is a funny coincidence. The homepage demo uses Curie, of of the faster GPT models but of lower quality. We think that different tradeoffs make more sense at different times. For instance, some applications come with long pauses for thinking (creative writing) and in these cases, it might be better to offer a slightly slower but more capable model. There is also a cost tradeoff for lowering the latency. The more aggressive you are with sending requests to OpenAI, the more likely it is for requests to be leapfrogged (the user keeps typing before the request is received). In the future, there might be smarter ways to reuse the computations on text prefixes using a custom model, but at the moment the latency does not seem to be the bottleneck for users.
If you try out the product in your own app, definitely leave a review with your thoughts!
Recommend
About Joyk
Aggregate valuable and interesting links.
Joyk means Joy of geeK