Banner-GPT

Hi everyone! Banner-GPT is an AI agent that can auto-generate banner images for any blog post or tweet. The premise is pretty simple. You paste your text into Banner-GPT, it reads and understands your content and then generates a banner image that best complements your writing. Under the hood, the raw post is sent to GPT3.5 with a few prompts to help it describe a visual that illustrates the main ideas in the text. The description is then sent as a prompt to Stable Diffusion XL. Finally the title is inferred from the content and framed on the image with a nice gradient overlay and turned into a downloadable image.

I am a humble person looking for a simple tool, and this is so perfect. I absolutely loved how it instantly came up with the best possible high-resolution images in a blink. How long are you planning to keep it free, or is there a forever freeway for limited usage? Have you decided on the pricing levels?

Really like your app. I tried using an article from Hacknews and the results of generating from it were very good. When the API be opened?

Love it! This is such a big time saver for people like me! I either try to create some shitty banner myself or have to get designer which are both time consuming. Do you also generate logos?

Much needed for writers like me. I'm gonna use thumbnail for my next blog.

Yo @neilxm this looks awesome! I just tried the latest version to generate banner images for some of Distribute Aid's recent social media posts, and love the results: post 1: https://www.instagram.com/p/CzEm... post 2: https://www.instagram.com/p/CzB9... post 3: https://www.instagram.com/p/CxEC... results: https://drive.google.com/drive/f... It would be cool if: 1. The images were available at a publicly shareable URL, so I can send them to colleagues or partners to review. Seems like a good flywheel for y'all. 2. Some way to specify the title text, hashtags, watermark / logo, or nothing (just the image). 3. Some sort of basic design settings. Color palette, star preferred styles, maybe upload a few general references images or some additional text context to include for certain types of posts? 4. A reach, but if there was an API available I'd love to integrate this with Strapi or our WikiJS setup, so that we can autogenerate images when content is added or updated. As for your questions: 1. I'd want to generate banners and stock images for our knowledge-sharing guides, blog posts, and social media posts. I think it'd also be a great way to come up with stylized profiles for team members who prefer their photos aren't posted online due to security concerns. it'd also be amazing to use this to tell a timeline-driven "supply-chain" story. there are often tons of great visuals that are easier / better to represent artistically instead of via poor-quality unprofessional photos (volunteers loading a container, a ship at sea carrying the aid, a truck driver in a foreign country, volunteers unloading in a war zone, distribution, a family with the items they need to survive another day). 2. It'd be great to upload a reference image that influences the style or content. When people are ffeatured, there needs to be a way to tailor the output for gender / race / age / hair color / cloting style. For example, it created a great banner image for post #2, but our featured volunteer Lizzie is white. Ultimately I'd want to configure most of the settings at an admin level, although possibly allow our content team to override them. This lets me setup templates while still allowing a human to be in charge, which paves the way for automatic banner generation based on CMS content, w/ 1-click approval from a human in the CMS before publishing.

It’s convenient ！very useful