Using the Figma MCP, it’s like saying, we have a car now, but I really want to go on a horse.
CODE
-
NumPy’s Simple File I/O API and PyTorch Tensor Limitations
By
–
Numpy’s fromfile()/tofile() make me smile — no typing or byte order metadata, they just write/read the raw bytes. A joyful API! I wish PyTorch had them natively on tensors, so I wasn’t doing so many tensor.cpu().numpy() operations.
-
DeepSeek V3.2 Sparse Attention Preparation for V4 Support
By
–
DeepSeek V 3.2 is basically just a heads up for inference tools devs to implement their new sparse attention mechanism so they can add V4 support on day 1
-
API Standards and Implementation Challenges in AI Development
By
–
I agree with you, I think that's why this isn't happening yet – plus there are still new features to be invented, a standard could slow that down But as someone who implements against these APIs I can dream!
-
Photo AI Hyper Realism Feature Deployed to Subscribers
By
–
✨ Okay I ended up calling it [ ⚡️ HYPER REALISM ]
— @levelsio (@levelsio) 30 septembre 2025
I deployed it to all Photo AI subscribers now!
⏰ 12 hours from me FAFO'ing to live feature (I also had to sleep 8 hours so actually 4 hours work)
You can enable it with one toggle in bottom left (made with Claude Code and a… https://t.co/W98mhepv9J pic.twitter.com/tk34ekZJgBOkay I ended up calling it [ HYPER REALISM ] I deployed it to all Photo AI subscribers now! 12 hours from me FAFO'ing to live feature (I also had to sleep 8 hours so actually 4 hours work) You can enable it with one toggle in bottom left (made with Claude Code and a
-
Deeper reasoning across all domains
By
–
Reasoning gains across every domain. Finance professionals, lawyers, medical researchers, STEM academics all tested it independently. Universal feedback: "dramatically better domain knowledge and reasoning" than Opus 4.1. Not faster answers. Deeper thinking.
-
Sonnet 4.5 tops SWE-bench leaderboard
By
–
SWE-bench Verified: state-of-the-art. This benchmark tests real software engineering. Actual GitHub issues that require multi-file edits, testing, dependency management. Sonnet 4.5 leads the leaderboard. It maintains focus for 30+ hours on complex tasks.
-
Anthropic’s Sonnet 4.5: A Coding Breakthrough
By
–
Anthropic dropped Sonnet 4.5 today. It's the best coding model in the world. Same price as Sonnet 4. But the capabilities gap is massive. This isn't incremental. It's a different category of reasoning.
-
Three.js 3D Environment with Camera Controls
By
–
2. Three.js 3D environment
— God of Prompt (@godofprompt) 30 septembre 2025
Requested an interactive 3D site with camera controls.
Claude generated a rotating, zoomable cube with ambient lighting and responsive canvas.
That's production-grade Three.js from a single prompt. Most devs still copy-paste examples for this. pic.twitter.com/q73JEZdeXR2. Three.js 3D environment Requested an interactive 3D site with camera controls. Claude generated a rotating, zoomable cube with ambient lighting and responsive canvas. That's production-grade Three.js from a single prompt. Most devs still copy-paste examples for this.
-
Claude 4.5 Sonnet: 3 Quick Projects
By
–
Claude 4.5 Sonnet is insanely powerful 🤯
— God of Prompt (@godofprompt) 30 septembre 2025
Here are 3 projects I made in it in few seconds:
1. Retro Snake in pure JavaScript pic.twitter.com/VTkTERWuTmClaude 4.5 Sonnet is insanely powerful Here are 3 projects I made in it in few seconds: 1. Retro Snake in pure JavaScript