Seedance 2.5 is now available on Open Design.
Native 30-second generation, up to 50 multimodal references, and far more precise control — the biggest leap yet for AI video.
Kimi K3 is now available on Open Design.
@Kimi_Moonshot 's latest model pairs powerful frontend capabilities with exceptional design taste.
Try it now ↓
Me: “Someone made a skill that lets Codex create breathtakingly beautiful Slides
It’s called Codex Slides, and it’s completely free and open-source.
You can throw it a GitHub repo, PDF, research document, or even a messy idea, and Codex will research the topic, create the outline, choose a visual style, and turn everything into a full presentation.
Every slide is generated as a complete visual image, I tested it out myself and ten slides took about four to five minutes to generate. You can then make changes, add speaker notes, and export the deck as a PowerPoint or PDF.
And this is where Codex has a massive advantage over Claude.
This workflow relies heavily on native image generation. Anthropic itself says Claude still can’t generate photos or illustrations, while Codex has ChatGPT’s image generator built directly into it. And in my opinion, ChatGPT is currently the best image generator in the world.
So Codex can now take almost anything and turn it into a genuinely good-looking presentation.
If you want the link comment "GPT" and I'll send it over”
Hey @elonmusk ! To celebrate Grok 4.5 going full Opus 4.8 and passing GPT-5.6, we’re doing 51% off Open Design Cloud Max for 3 days. 🎉
Now including:
Grok 4.5 · GPT-5.6 · Claude Fable 5 · GLM-5.2 · Muse Spark 1.1
→
To celebrate GPT-5.6 having its Claude Opus moment, we’re doing 51% off Open Design Cloud Max for 3 days. 🎉
Now including:
GPT-5.6 · Grok 4.5 · Claude Fable 5 · GLM-5.2 · Muse Spark 1.1
→
GPT-5.6 is now supported in Open Design, including Sol, Terra and Luna.
No Codex Pro required. Open Open Design Cloud, pick any model you like — GPT-5.6 included — and switch anytime, context intact.
Codex CLI is available too.
Seedream 5.0 Pro API is now available on Open Design.
Bring your own Volcengine (Ark) API key, and generate production-ready visuals with Seedream 5.0 Pro right inside your design workflow.
Introducing motion-anything — an open-source alternative to Figma Motion, just landed in the Open Design family.
Just describe the feeling, your AI ships the animation.
403 recipes · 230 Skills · 59 design systems · 8+ code agents + BYOK
Check more info below⬇️
The bottleneck isn’t having a brand.
It’s turning it into every report, deck, and customer story your team needs.
Ikigai One used Open Design for exactly that — and called Open Design “invaluable” after one month.
Read:
70K+ GitHub stars in 57 days.
Since launching on Apr 28, we’ve reached:
#1# TypeScript repo of the day/week
#2# repo of the day across all languages
#3# repo of the week across all languages
Any model. Any agent. Open design for everyone.
These two Seedance 2.5 demos are especially exciting:
Native 4K cinematic realism.
Refined watercolor-and-gold-foil promo visuals.
Seedance 2.5 is coming soon to Open Design, along with the latest and most complete Seedance 2.5 prompt library.
Seedance 2.5 is coming soon to Open Design.
30s video generation.
Native 4K.
Up to 50 multimodal references.
3D white-model support.
Licensed film IP creation and revenue sharing.
See more demos below 👇
We tried GLM 5.2 inside Open Design’s built-in AMR.
Input:
- a reference website link
- one simple prompt
Output:
- a near-perfect website recreation
No detailed spec. No long prompt engineering. Just reference-to-web, directly inside Open Design.
We hit 68,000 🌟 GitHub stars in 7 weeks.
We replaced human PR review with a loop. Median time-to-merge dropped to 3.4 hours, ~30 PRs merged a day.
Here's how we built it 🧵