Ali M

[Tripo @tripoai]

Introducing Project Eden, a world model research preview from @VASTAIResearch

Project Eden is a persistent, multiplayer world model that fundamentally breaks from existing paradigms by decoupling the underlying world state from visual rendering.

Instead of treating the world as a sequence of transient frames, Eden treats it as a structured, evolving environment that runs continuously, can be modified by user actions, and can be consistently observed from any viewpoint.

[Claude @claudeai]

Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors.

Available today at the same price.

[Sergey Karayev @sergeykarayev]

I don’t understand how this can be 2.6B params

[Tengfei Wang @DylanTFWang]

⚡️After weeks of hard work, we're thrilled to fully open-source HY World 2.0 today -- full inference code and all models!

Build, explore, and create your own interactive worlds with us.👇
https://t.co/cfLustgjE5

[Overworld @overworld_ai]

Today we’re releasing Waypoint-1.5.

An update to our real-time diffusion world model designed to run interactively on consumer hardware.

[AK @_akhaliq]

INSPATIO-WORLD

A Real-Time 4D World Simulator via Spatiotemporal Autoregressive Modeling

paper: https://t.co/8pRBe73t6B

[AK @_akhaliq]

MultiGen

Level-Design for Editable Multiplayer Worlds in Diffusion Game Engines

paper: https://t.co/UT01uUYBxj

[Skywork @Skywork_ai]

Skywork Matrix-Game 3.0 is here! FULLY OPEN SOURCE!
Real-Time and Streaming Interactive World Model with Long-Horizon Memory

- Fully open source: code, model, and technical report
- 720p @ 40FPS with a 5B model
- Minute-long memory consistency
- Trained on Unreal Engine + AAA games + real-world data
- Scales up to 28B MoE for quality, dynamics, and generalization

Homepage 👉 https://t.co/DVMV6UdINu
Code 👉 https://t.co/VwLWTpEsbq
Model 👉 https://t.co/JMSCqdtydH
Tech report 👉 https://t.co/GEs1WrUNhe

Create. Explore. Play. With Matrix-Game 3.0

[AK @_akhaliq]

AnchorWeave

World-Consistent Video Generation with Retrieved Local Spatial Memories

paper: https://t.co/aiKcOIYpVC

[Javi Lopez ⛩️ @javilopen]

⚡ Google Genie 3 but OPEN SOURCE

Not even 48h later and the Chinese did it again: they just dropped a free real-time playable world generator.

- LingBot-World
- Built on Alibaba's Wan2.2
- REAL-TIME interaction at 16fps

100% open source 🧵

[Emm | scenario.com @emmanuel_2m]

If you’re working in 3D and haven’t tried Sparc3D 2.0 yet,

especially the "1536 Pro“ mode,

... you’re seriously missing out.

👉 Here’s where to access it, along with 25+ other generative 3D models: https://t.co/Xj5tZyLiLK

[AK @_akhaliq]

Microsoft just released VibeVoice-Realtime-0.5B
https://t.co/uIqp5yrZM6

[Tanishq Mathew Abraham, Ph.D. @iScienceLuvr]

WorldGen: From Text to Traversable and Interactive 3D Worlds

"We introduce WorldGen, a system that enables the automatic creation of large-scale, interactive
3D worlds directly from text prompts."

Four stages: scene planning, scene reconstruction, scene decomposition, scene enhancement

[Gabriele Romagnoli @GabRoXR]

AI Mocap tools are evolving at breakneck speed... and I could find more than 10 all compiled and compared in a list you can get by subscribing to the XR AI Spotlight newsletter (link on top of my profile).

This is @getcartwheel and the stepping on that box was one of the most solid I saw. There is a bit of floatiness in the feet and they are not perfectly grounded, but since I tested the tool, the team has pushed massive improvement.

What AI mocap tool do you use in your workflow?

Let me know in the comments

[Tencent Hy @TencentHunyuan]

The training code for Hunyuan World 1.1 (WorldMirror) is released now!🔥🔥🔥

This release provides researchers and developers the full stack for customization and fine-tuning:
📷 Your video to 3D worlds in 1 second.
🪄ANY input (image, video, 3D prior) to ANY output (3DGS, depth, cameras, normal, points)

Start customizing your model and exploring the demo!
Online demo: https://t.co/gMQwz0vf35
Code: https://t.co/tOjM5BAr9o

[Deedy @deedydas]

Two 23 year old Indians just dropped the #2 open-weight AI voice model in the world, trained purely on free credits!

Maya1 is #20 globally, better than even Google's best. 3B params, runs on one GPU and does 20+ emotions with < 100ms latency

You can just do things.

[Tencent Hy @TencentHunyuan]

Today, we are open-sourcing Hunyuan World 1.1 (WorldMirror), a universal feed-forward 3D reconstruction model. 🚀🚀🚀

While our previously released Hunyuan World 1.0 (open-sourced, lite version deployable on consumer GPUs) focused on generating 3D worlds from text or single-view images, Hunyuan World 1.1 significantly expands the input scope by unlocking video-to-3D and multi-view-to-3D world creation.

Highlights:
🔹Any Input, Maximized Flexibility and Fidelity: Flexibly integrates diverse geometric priors (camera poses, intrinsics, depth maps) to resolve structural ambiguities and ensure geometrically consistent 3D outputs.
🔹Any Output, SOTA Results：This elegant architecture simultaneously generates multiple 3D representations: dense point clouds, multi-view depth maps, camera parameters, surface normals, and 3D Gaussian Splattings.
🔹Single-GPU & Fast Inference: As an all-in-one, feed-forward model, Hunyuan World 1.1 runs on a single GPU and delivers all 3D attributes in a single forward pass, within seconds.

🌐Project Page: https://t.co/bbFzdTXUdU
🔗Github：https://t.co/tOjM5BAr9o
🤗Hugging Face：https://t.co/LwnlHiFFLX
✨Demo: https://t.co/gMQwz0vf35
📄Technical Report: https://t.co/qIGXB1znU0

[🚨 AI News | TestingCatalog @testingcatalog]

BREAKING 🚨: Google started working on a Genie 3 experiment where users will be able to generate their own worlds to explore.

Genie 3 is an AI World Model from Google announced earlier this year.

[AK @_akhaliq]

StreamingVLM

Real-Time Understanding for Infinite Video Streams

[Robert Youssef @rryssf]

RIP fine-tuning ☠️

This new Stanford paper just killed it.

It’s called 'Agentic Context Engineering (ACE)' and it proves you can make models smarter without touching a single weight.

Instead of retraining, ACE evolves the context itself.

The model writes, reflects, and edits its own prompt over and over until it becomes a self-improving system.

Think of it like the model keeping a growing notebook of what works.
Each failure becomes a strategy. Each success becomes a rule.

The results are absurd:

+10.6% better than GPT-4–powered agents on AppWorld.
+8.6% on finance reasoning.
86.9% lower cost and latency.
No labels. Just feedback.

Everyone’s been obsessed with “short, clean” prompts.

ACE flips that. It builds long, detailed evolving playbooks that never forget. And it works because LLMs don’t want simplicity, they want *context density.

If this scales, the next generation of AI won’t be “fine-tuned.”
It’ll be self-tuned.

We’re entering the era of living prompts.