You're right, the market is flooded with simple "button grid" apps. That saturation is actually why I built FlowPDF.
I didn't want another list of basic tools; I wanted to chain them. I built a node-based editor so you can create actual pipelines (Merge -> OCR -> Filter -> Compress) rather than just doing one-off tasks.
I think that's the only way to actually add value over the 50 other "Hello World" clones.
I've been working on exactly this with FlowPDF (https://www.flowpdf.app/). It was a huge pain to find a tool that didn't rely on Google APIs.
I integrated a local translation model. It downloads the model to your browser cache on the first run (so it's a bit heavy initially), but after that, you can OCR and translate documents 100% offline without uploading anything.
I actually built a tool called FlowPDF - https://www.flowpdf.app/ - specifically to solve this "logic" problem. I found that single-button tools were too limiting for things like "Merge -> Check Size -> Compress".
It uses a node-based graph so you can literally string those steps together (e.g., Merge files -> Split out page 35 -> Compress the rest) and save it as a repeatable workflow.
Codex just seems to have a much bigger context and doesn't chew up tokens as readily as Claude. It seems to be able to do a much wider and broader range of things, accepting much wider and broader instructions and implementing them perfectly.Whereas Claude would struggle.
I know that a lot of content on the web is rehashed itself, but many creators actually review products themselves, speak from first hard experience, research, etc.
The original knowledge has to come from someone human.
How can Google promote its "1 trillion facts" database of scrapped content and just use it to show answers to searchers alongside ads.
Seems like they are doing this without regard for much anything else because OpenAI is a new entrant and about to threaten their search empire?