News for hackers. Stuff that matters.

Applemoi

I sometimes feel like its a rough transcript of the audio or similar. Because not only is the response apparently natively audio, but I've often seen the text adhere to the audio only as a best effort rather than a full on accurate representation of the audio.
Needed something to do this long weekend and I have no clue where I got this idea from, but the concept is simple - vote between two colors at a time, collate the results, and you have a crowd sourced ranking of the best to worst colors! Let me know your thoughts or how the website could be improved further. And hope you have fun voting!
Votes are rolling in! Exciting to see the leaderboard update.

Show HN: Multimodal Gemini YouTube Chat – Chat with full visual understanding

https://youtubechat.agpallav.com/

Chat with a YouTube video using Gemini’s Multimodal understandingChat with any YouTube video using Gemini’s multimodal API - this is the first YouTube AI Chat that brings _visual_ understanding to the chat.Current websites for chatting with a YouTube video usually just use the transcript. I used Gemini’s multimodal capabilities to allow you to chat with FULL audio and visual understanding.

( Read More... | no comments | on HN )
I highly recommend using GPT-4 Turbo and DALLE-3 from the API, especially if you just want to try it out. It’s not free but the API doesn’t require an upfront flat fee and doesn’t have rate limits etc!

There’s some options for desktop like TypingMind.

For iOS I wanted to suggest my app called Pal Chat that lets you use your own API Key to chat with OpenAI and other models, and it’s completely free to download.

The app also has no backend and takes absolutely no usage data. You can verify this by going to the app privacy report in iOS settings.

You have conversation history and you can use your own API Key on this app that supports GPT-4 Turbo, Claude 2.1, Gemini Pro and DALL-E 3. I’m adding mistral soon! https://apps.apple.com/us/app/pal-ai-chat-client/id644754508...

Show HN: Native Bard iOS App

https://apps.apple.com/us/app/pal-ai-chat-client/id6447545085

This iOS app supports Google's new Gemini Pro model - making it a pseudo, arguable better, interface for interacting with Bard/Gemini Pro. It's powered by your own API Key so there's no backend involved.The app also Supports GPT-4 Turbo, DALL-E 3, Claude 2.1 and any OpenAI based model (eg. OpenRouter)

( Read More... | no comments | on HN )

Show HN: Native GPT-4 and Gemini Pro iOS Client

https://apps.apple.com/us/app/pal-ai-chat-client/id6447545085

I wanted a way to interact with GPT-4 Turbo and DALL·E 3 without having to pay a minimum of $20 for ChatGPT Plus - so I made this iOS wrapper client to achieve just that.It's powered by your own API Key so there's no backend, and it also supports on-device conversation history.It started off with GPT-4 but now also support Claude 2.1 and Gemini Pro with their respective API Keys.Another very...

( Read More... | no comments | 2 points | on HN )

Show HN: DALL-E 3 and GPT-4 Vision Recursive Image Generator

https://image-recursor.web.app

It's like Gartic Phone for images, powered by AI. I created an automated DALL·E 3 and GPT-4 Vision Image Recursor.Simply provide a starting prompt and see how AI iterates on it. It's perfect for the 'Make it more' trend.You'll need to provide your own OpenAI API Key, but the website is open source! I've linked to the GitHub on the website.Try it out: https://image-recursor.web.app/

( Read More... | no comments | on HN )
Hey HN! Excited to share my free iOS app - Pal!

At a high level, the app works by using the ChatGPT API and your own API Key. This allows you to use GPT-3.5 and GPT-4 through the API rather than having to pay a flat $20/month for ChatGPT Plus!

I started this as a side project to use myself - but some friends found it so useful I decided to just drop it as a free app on the App Store.

I eventually added some nice to have features such as: - API Key access to GPT-3.5, GPT-4, DALL•E (and Google PaLM!) - Conversation History Support (New!) - Ability to edit both the question and the AI response (Famously missing in the official app) - Full syntax highlighting support - Saving multiple prompts - Saving multiple system messages - Ability to adjust the temperature - See the token length and cost of the conversation

Hope you find the app useful!

And of course, the app collects no data! You can verify this by opening Settings and going to Privacy>App Privacy Report.

Please remember to get an API Key from OpenAI!

tl;dr Use an API Key to pay-per-use rather than $20/month for GPT-4 access