addy999

4 ms·

1.
Show HN: NativeBlend – Text to fully editable 3D Models that don't suck(native-blend-app.vercel.app)
2 points by addy999 11mo ago|1 comments
2.
by addy999 1y ago
In my testing I tried Gemini 2.5 Pro, Qwen 3 Coder 480B and Opus
3.
Building Effective Text-to-3D AI Agents: A Hybrid Architecture Approach(addy.rocks)
20 points by addy999 1y ago|3 comments
4.
I built a Context7 alternative that costs 40% less with similar code quality
1 points by addy999 1y ago|0 comments
5.
LLMs can now generate 3D models (example)(i.imgur.com)
3 points by addy999 1y ago|1 comments
6.
by addy999 1y ago
Thanks to tools like Blender and MCP servers, models from Gemini to Qwen can be used to create fully usable models. Having LLMs generate code to build a model by using tools like an artist would drastically increases quality over models pro
7.
by addy999 2y ago
I'm building an AI web agent to fetch detailed information. Looks up and browses sites just like you. Have just over 500 developers signed up and releasing our client library soon. https://www.onequery.app
8.
Web agents can simplify web browsing(onequery.app)
2 points by addy999 2y ago|0 comments
9.
by addy999 2y ago
I find v0.dev is remarkably good at this. But deploying straight from there is not frictionless atm.
10.
I liked Microsoft's OmniParser model so wrapped it in an API(github.com)
1 points by addy999 2y ago|0 comments
11.
by addy999 2y ago
Your approach makes a lot of sense. Especially for "old school" websites with simple Server Side rendered pages, tables are a treasure of data. Wikipedia tables would be a great use of your tool too. I was thinking more on the lin
12.
Ask HN: How do you handle repetitive research tasks in your workflow?
1 points by addy999 2y ago|2 comments