AI & Automation Builder 🤖 Building agents, workflows & smart systems. 🛠️ Open-source, IoT & LLMs. Sharing what I build, test & learn. DM for collabs.
📩DM for collabs
Joined November 2011
- Tweets106K
- Following72.8K
- Followers11K
- Likes35.4K
Pinned Tweet
"Honestly, this is a pretty good time to give AI video creation a go. 👀
Dreamina handles the visuals, CapCut PC handles the edits, and the combo starts at just $6.29. Ends Oct 30!
#CapCutDidThat #Dreamina #Smallbusinessmarketing #CapCutSale"
"Turning a portfolio into a short introduction feels worth trying with Dreamina’s first month at $0.29. CapCut’s live Black Friday sale offers 60% off Pro and 40% off Ultra Oct 20–30, desktop only.
Use Dreamina for the visual framing, then bring those assets and your actual work samples into CapCut PC. Start with the strongest example, add brief project labels, and keep the transitions simple. Finish on what you do and how to get in touch.
At regular pricing, Dreamina and CapCut Pro are $15/month each. Purchase the subscriptions independently through their own offer links; both deals end October 30.
Give prospective clients an easy first look at your work.
CapCut: tinyurl.com/2v23rn22
Dreamina: tinyurl.com/5xbtv4x6
#CapCutDidThat #Dreamina #Smallbusinessmarketing #CapCutSale"
Me seeing another Tesla Roadster post and immediately imagining a cinematic edit.City lights, dramatic scenery, a strong visual mood — now I just need to figure out how I want to tell the story. The CapCut plugin is coming soon to ChatGPT, where you can describe your idea and work with CapCut Web to explore it.Early Access is currently available through ChatGPT desktop (PC only).
#CapCutWeb #CapCutDidThat #ChatGPT #CapCutWorld
Try
tinyurl.com/2xeju4vv.
Zuar⚡ retweeted
Our agent scientist, Agent S4, just got an open 27B model to beat GPT-5.5 and Claude Opus 4.8 on OSWorld V2, with the model's weights left exactly as they were.
With the interfaces S4 evolved, Qwen3.8-27B completes 21.3% of the 108 tasks, the highest binary success of any open-source computer-use agent and 3.8x the same model with fixed interfaces.
It costs $1.10 per task, against $25.50 for GPT-5.5 and $72.40 for Opus 4.8 💸
Full write-up: simular.ai/articles/agent-s4…
Smaller models, smarter systems. 🔥
Agent S4 helped Qwen3.8-27B achieve a 21.3% success rate on OSWorld V2 without changing the model's weights.
At just $1.10 per task, this is a strong example of how better interfaces can unlock more potential from open-source AI.
Full details: simular.ai/articles/agent-sc…
Our agent scientist, Agent S4, just got an open 27B model to beat GPT-5.5 and Claude Opus 4.8 on OSWorld V2, with the model's weights left exactly as they were.
With the interfaces S4 evolved, Qwen3.8-27B completes 21.3% of the 108 tasks, the highest binary success of any open-source computer-use agent and 3.8x the same model with fixed interfaces.
It costs $1.10 per task, against $25.50 for GPT-5.5 and $72.40 for Opus 4.8 💸
Full write-up: simular.ai/articles/agent-s4…
Big news for the design community! ⚡
Ideogram has introduced its first Vector model, bringing AI-powered vector creation to a new level. Create scalable graphics with precise typography and gradients, ready for further editing.
🔗 Explore it here: ideogram.ai/models/vector
🤝 Paid partnership
Today, we’re introducing Ideogram's first vector model.
Generate clean, editable SVGs with smooth gradients and sharp typography from a text prompt, or turn any image into a detailed, scalable vector.
Live in Ideogram and the API: ideogram.ai/models/vector
Zuar⚡ retweeted
Today, we’re introducing Ideogram's first vector model.
Generate clean, editable SVGs with smooth gradients and sharp typography from a text prompt, or turn any image into a detailed, scalable vector.
Live in Ideogram and the API: ideogram.ai/models/vector
Zuar⚡ retweeted
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
Open-source AI is moving fast! 🔥 Ornith-1.5-122B brings power, speed, and efficiency together.
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
Zuar⚡ retweeted
Open weights for @StepFun_ai’s Step 5 Preview are live. 600B total, 27B active, 1M context, and No. 2 on @OpenRouter’s weekly chart. We read the tech blog so you don’t have to
AI content, real creators, and licensed likenesses. 🚀
@mirrormirror_ai is making AI creation more ethical and accessible!
AI slop isn't just tacky and uninspired. It can get you sued.
Generating fake AI faces that resemble real people, or using someone's likeness without permission, is creating new legal and reputational risks for brands.
So we built @mirrormirror_ai as part of a16z @speedrun, a marketplace where brands can digitally book real models and creators to generate AI content using their likeness, with their permission.
Brands browse and book identity-verified talent, agree on usage rights, and generate content within those terms. Every approved image comes with a signed license and embedded provenance.
We're seeing brands produce lookbooks and campaign visuals up to 10x faster, with up to 90% cost savings. Meanwhile, models and creators can earn additional income from their digital likeness without having to show up for another shoot.
Everyone is racing to make AI generation faster, cheaper, and more realistic. But there's still a huge missing piece: how do you license the rights to the people appearing in AI-generated content at scale?
That's why I'm excited about what we're building. Not just another AI generator, but the licensing and rights infrastructure for AI-generated content.
mirrormirrorai.com/
Zuar⚡ retweeted
AI slop isn't just tacky and uninspired. It can get you sued.
Generating fake AI faces that resemble real people, or using someone's likeness without permission, is creating new legal and reputational risks for brands.
So we built @mirrormirror_ai as part of a16z @speedrun, a marketplace where brands can digitally book real models and creators to generate AI content using their likeness, with their permission.
Brands browse and book identity-verified talent, agree on usage rights, and generate content within those terms. Every approved image comes with a signed license and embedded provenance.
We're seeing brands produce lookbooks and campaign visuals up to 10x faster, with up to 90% cost savings. Meanwhile, models and creators can earn additional income from their digital likeness without having to show up for another shoot.
Everyone is racing to make AI generation faster, cheaper, and more realistic. But there's still a huge missing piece: how do you license the rights to the people appearing in AI-generated content at scale?
That's why I'm excited about what we're building. Not just another AI generator, but the licensing and rights infrastructure for AI-generated content.
mirrormirrorai.com/
Zuar⚡ retweeted
Your team shouldn't be trapped in separate windows with agents and tools.
With Kylon, your Claude Code, Codex and your people work in the same room where they talk to each other, build and ship real work.
The company harness brings in the tools your team spent months learning, so people get to focus on more interesting work.
