News for hackers. Stuff that matters.

qzcanoe

Pony Alpha is a stealth foundation model launched on Feb 6, 2026 via OpenRouter/Kilo. It offers a 200 k token context window and up to 131 k token outputs, with $0 listing. Key modes include coding, reasoning and role‑play. Our page explains what the model is, what people are searching for and how to test it safely.

Show HN: A fast, simple playground for Trellis 2 image generation

https://trellis2.com/

Trellis 2 is a new image generation model, but I found it surprisingly hard to quickly try, compare prompts, or understand what it’s good at without digging through docs or code.So I built a small, fast site that focuses on one thing: letting people experiment with Trellis 2 through real prompts and instant results.It’s intentionally simple:* No account required * Prompt-in, image-out *...

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

Show HN: Wan-Animate – AI video character animation and replacement based

https://wan-animate.com/

I built Wan-Animate, an AI video character animation and replacement tool based on the Wan 2.2 model.It focuses on two common video creation needs:Move Mode: animate a static character image to follow the motion in a reference videoMix Mode: replace the person in a video with an uploaded character imageUnder the hood it uses skeleton-driven motion control and intelligent relighting, so the...

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

Show HN: FocusFour – Eisenhower Matrix on Apple Reminders

https://www.focusfour.app/

I built FocusFour because I was busy all day but still missing the things that actually mattered.FocusFour is a task manager based on the Eisenhower Matrix, built directly on top of Apple Reminders. Recently I added lightweight AI to help identify and surface tasks that are likely important but getting ignored.What it does:* Four-quadrant (important vs. urgent) task view * Uses AI to highlight...

( Read More... | no comments | on HN )
Re:Show HN: PhotoToVideoAI – AI photo to video generator on Monday December 15, 2025 @08:45AM (#46274446)
Thank you!

ComfyUI Cloud is very powerful and flexible, especially if you want full control over nodes and workflows. PhotoToVideoAI is aimed at a different audience: people who just want to go from a single photo + a simple prompt to a usable video as fast as possible.

We focus on:

* a very low learning curve (no workflow setup) * consistent presets for photo-to-video use cases * fast generation with predictable results

In short, it’s less about maximum flexibility, more about speed and simplicity. Some users actually use both, depending on the task.

Re:Show HN: PhotoToVideoAI – AI photo to video generator on Monday December 15, 2025 @08:45AM (#46274440)
Thanks for the kind words!

We currently offer a no-questions-asked refund within 7 days if the product doesn’t meet your expectations. Since generation costs are involved, we try to keep it fair for both users and us, but if something goes wrong or the results aren’t usable, I’m always happy to help or refund manually.

Happy to clarify anything if needed

Show HN: PhotoToVideoAI – AI photo to video generator

https://phototovideoai.org

PhotoToVideoAI is an AI-powered tool that turns your photos into dynamic videos. Upload a photo and a prompt, and in about 30 seconds you'll get a high-quality video in resolutions up to 1080p and durations of 5 or 10 seconds. It’s designed for content creators, marketers, and photographers, Feedback is welcome!

( Read More... | 4 comments | 3 points | on HN )

Show HN: NBPro – NanoBanana Pro prompt library (100 prompts)

https://nbpro.org/

NBPro is a dedicated NanoBanana Pro prompt library built for creators. Explore over 100 curated and proven prompts across real commercial use cases like product shots, ads, social visuals, and more – all optimized for NanoBanana Pro.Core value: - Save time on searching and trial‑and‑error - Recreate real‑world effects instantly using one‑click prompts - Learn prompt structure and best...

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

Show HN: EraseVideo – a Free Mac app removes Sora video watermark in 1 minute

https://erasevideo.app

EraseVideo, a macOS application designed for AI video creators who want a simple, privacy-first solution:Drag in a video → automatically detect & remove the watermark → export a clean version — all processed locally on your Mac, no upload, no signup required.Background & problemAs AI-video tools (e.g., Sora) became more commonplace, creators found themselves hindered by persistent...

( Read More... | no comments | on HN )
Re:Free Normal Map Online Generator on Friday September 05, 2025 @09:03AM (#45138065)
When working on 3D projects, I often needed to generate normal maps quickly. Traditional methods required opening heavy software or installing complex plugins. Many developers, artists, and hobbyists really just want a simple, instant, online tool. That’s why I built Normal Map Online.

What it does

Normal Map Online is a free online normal map generator that lets you: • Upload a texture or height map. • Instantly generate a high-quality normal map. • Preview the result in real time with a 3D viewer. • Download the map as PNG/JPG for direct use in Unity, Unreal, Blender, and other engines.

How I built it

I built the tool using HTML, CSS, JavaScript, and WebGL (Three.js). The interface is styled with TailwindCSS for a modern, minimal look. The workflow is designed around “Upload → Adjust → Download,” so users can get results in just a few seconds.

Challenges I ran into • WebGL rendering consistency: different browsers and devices handle lighting and textures differently, which required extra tuning. • Balancing simplicity and completeness: keeping the interface clean for beginners while still offering advanced features for power users. • SEO and performance: making the tool lightweight and fast while also discoverable by search engines.

Accomplishments that I’m proud of • Created a truly instant web tool that works without any installation. • Designed a modern, minimalist UI that makes complex features approachable. • Helped developers and creators save time and improve their workflow.

What I learned • UI/UX details make or break usability — even small things like slider placement and hint text matter. • When launching niche tools, SEO strategy and product positioning are just as important as the core functionality. • How to handle WebGL cross-browser quirks and debug 3D rendering issues.

What’s next for Normal Map Online • Batch processing: generate multiple maps at once. • Higher-resolution exports for professional use cases. • API access for developers to integrate into pipelines. • More learning resources and tutorials to help 3D beginners get started quickly.

Hey Samstave, thanks for the feedback! Glad you find the tool useful.

Actually, you can already edit the Markdown output directly in the left window, and the preview on the right will update in real-time.

I'm actively working on adding more features, and tabs, history, and batch bookmark importing are definitely on the roadmap!

Thanks again for trying it out and sharing your thoughts – it's super valuable as I continue to improve the tool.

Show HN: LinkToMarkdown – Extract web content to clean Markdown in one click

https://linktomarkdown.com/

Hi!I recently created a little tool called LinkToMarkdown (<https://linktomarkdown.com/>](https://linktomarkdown.com/)). It extracts web content and generates clean Markdown text with just one click.To be honest, I developed this tool purely out of personal need. I often want to save articles from the web, but I'm always frustrated by ads, pop-ups, and all sorts of clutter. One day, I'd...

( Read More... | 3 comments | 2 points | on HN )