Developer | AI News & testing | Startups & SaaS | DM for Collabs ☄
Joined April 2025
- Tweets17.2K
- Following672
- Followers3.8K
- Likes58.8K
Adit_Yah ☄️ retweeted
Holy… Kimi K3.1 is insane 🤯
It created an entire launch video for Kimi K3.1 using Three.js in just 5 minutes for $1.10.
WHAT A COMEBACK! 🔥
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model.
🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal.
🔹 Full-project coding: games in Unreal, Unity & Godot.
🔹 Excels at parametric 3D modeling in Blender & CAD.
🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%.
🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks.
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Open weights by Nov 4, 2026.
🔗 API: platform.kimi.ai
🔗 Tech blog: kimi.ai/blog/kimi-k3-1-pro
Adit_Yah ☄️ retweeted
MiniMax M3.1 Flash is actually insane.
I asked it to build this whale breach scene using Claude Code as the harness, and it cost just $0.25. That's roughly 8x cheaper than Astra and Sol.
I didn't expect MiniMax to be this good at 3D.
In my testing, it even beats GPT 6.1 Sol, and the result looks better than the others while costing less than DeepSeek Flash’s version.
What a comeback by MiniMax.
Adit_Yah ☄️ retweeted
Gemini 4 Argon is crazy ,
it just took 10 mins and less than 2% of my weekly quota to built the "Seveen wonder of worlds"
I tested it multiple times and every time the output is clean, consistent and impressive
Adit_Yah ☄️ retweeted
What if you could rent a piece of a creator’s X profile?
I built it.
Introducing BookMyProfile -a way for brands/founders to book:
→ My bio link
→ My cover/banner
→ A dedicated promo post
All with my audience stats and impression data visible before booking.
I’m turning my X profile into bookable ad space.
22M+ impressions in the last 3 months.
Now, brands can book a spot.
bookmyprofile.xyz
Adit_Yah ☄️ retweeted
4 Argon vs Opus 5.5 vs GPT 6 Astra vs Kimi K3
I asked Argon to generate this rocket scene.
Argon focuses more on detail, and I'm loving it. The result isn't as good as Opus or Kimi, but the realism is still there. and Its efficient.
I never expected this from Google.
Adit_Yah ☄️ retweeted
4 Argon vs Opus 5.5 vs GPT 6 Astra vs Kimi K3
I asked all four to build a 3D whale breaching scene.
Argon's level of detail is on another level, and it wins on photorealism. Opus and Kimi look more realistic than Astra.
But what a comeback from Google!! 🔥
Adit_Yah ☄️ retweeted
Build a 45s scene with STEP 5 PREVIEW
It took ~8 mins to build this rocket launch scene
Single html file , pure three.js, no external assets
it's is now live on OpenRouter and available in Kilo, Cline, Kilo Code, Hermes Agent, and OpenCode.
🤝 Paid partnership
Adit_Yah ☄️ retweeted
Step 5 Preview now exists where you build.
Now live on @OpenRouter. A week of free access is rolling out today across @opencode, @cline , @NousResearch, @kilocode, and more.
Flagship-tier intelligence for agentic and professional work, at substantially lower task cost.
Keep your workflow. Switch your model.
Try Step 5 Preview on OpenRouter: openrouter.ai/stepfun/step-5…
Adit_Yah ☄️ retweeted
If this gets integrated into Zoom or Teams, I genuinely don’t know how you’d tell it’s AI 😭
Griffin can see, hear, react and respond in real time, and it can even interrupt you or notice someone walking into the frame.
This is getting wild.
Introducing Griffin, the first model to pass the video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
It’s the first Human Interaction Model (HIM).
Readers added context they thought people might want to know
The 48% figure and "video Turing test" claim are from Tavus's own study of 54 one-minute calls, not independently verified or using a standard protocol. Griffin-Lite leads NVIDIA's VideoFDB benchmark on their public leaderboard.
cellcog.ai/blog/tavus-gri…
research.nvidia.com/labs/amri/proj…
tech-ish.com/2026/10/02/tav…
Adit_Yah ☄️ retweeted
Grok 4.7 is here.
It's a notable improvement over Grok 4.6 at the same price and speed.
Adit_Yah ☄️ retweeted
GPT-6 Luna may already be in A/B testing.
Some GPT-5.6 Luna requests are being served as GPT-6 Luna behind the scenes.
github.com/d4rken/clankermux…
Adit_Yah ☄️ retweeted
MiniMax Code is open source now.
I forked it and added a live cost panel right in the terminal, showing tokens, cache hits, and session cost in real time.
What I like is that MCode lets you connect MCP tools, use built-in or custom Skills, and switch between different model providers directly from the CLI.
You can actually extend the coding agent to fit your own workflow.
This is what open source should enable.
Adit_Yah ☄️ retweeted
This is Fake Opus 5.5 output
He stole the video from YouTube:
youtu.be/AlMLFkN79dA?si=SXP4…
Thanks for the info @Deepusleepy
This quoted post is unavailable.