AI in week 24.

CF

Claude Fable 5

Model · Anthropic

18

Anthropic's premium Claude model for demanding technical work in coding, finance and science.

G5

GLM 5.2

Model · Zhipu AI

10

Zhipu AI's GLM 5.2 large language model family, with Fast and max variants, available via Z.ai chat and API.

Anthropic model reported by creators to have no cyber guardrails and stronger exploit development than Opus 5.

SA

Siri AI

App · Apple

4

Apple's upgraded Siri assistant built on Apple Intelligence, with on-screen and cross-app context; not available in the EU.

Google's open-weights 26B diffusion language model that drafts blocks of text in parallel.

KK

Kimi K2.7 Code

Model · Moonshot AI

1

Moonshot AI's Kimi K2.7 coding model, offered through the Kimi Code agent (CLI, VS Code extension, desktop app).

NN

Nex N2

Model · Nex AGI

1

Open agentic model family from Nex AGI for computer use, coding and long-horizon tool tasks, in several sizes.

G3

Gemini 3.1 Pro

Model · Google

1

Google's Gemini 3.1 Pro multimodal large language model, available in the Gemini app and via the Gemini API.

I1

i1

Model · Princeton

1

Fully open 3B text-to-image diffusion model from Princeton with released weights, training code, and data pipeline.

S2

SCAIL 2

Model · Z.ai

1

Research framework for end-to-end controlled character animation and motion transfer from a driving video to a reference character.

MI

Millivid

Research

1

Research method for long-context video generation using hierarchical latents and coarse-to-fine rollout.

Research video generation model controlled by local or global force signals, with causal streaming generation.

G3

Gemini 3.5 Live Translate

Model · Google

2

Google's real-time speech translation model that preserves the speaker's voice, intonation and pacing across 70+ languages.

Open-source 2B-parameter zero-shot voice-cloning text-to-speech model from Rednote's hilab team.

AR

Arbor

Model · Stability AI

2

Research method adding typed 3D constraint meshes (hull, avoidance, touch) as control for text-to-3D generation on TRELLIS.2.

AN

AnchorWorld

Research

1

Research framework for first-person world simulation driven by 3D human motion and anchor-view scene customization.

MO

Moverse

Research

1

Research system that turns a single image into a 360° Gaussian scaffold and renders interactive video at 8 FPS.

FL

Flex4DHuman

Research

1

Multi-view video diffusion model that turns monocular or sparse videos of dynamic humans into dense multi-view videos for 4D Gaussian splats.

AW

Actionable World Representation

Research

1

Research method that builds controllable, articulated 3D object models from keypoints, point clouds or depth video.

ME

MeshFlow

Model · Meta

1

Meta and HKUST research method for fast artist-like 3D mesh generation using a MeshVAE and flow-based diffusion transformer.

SU

Surflo

Research

1

Research method that reconstructs 3D surfaces from a variable number of unposed images using flow matching.

WT

World Tracing

Model

1

Research model that predicts layered, pixel-aligned 3D point clouds, including hidden geometry, from one image or video.

GO

/goal

App · OpenAI

4

Codex command that keeps an agent working on a long-running goal until a verifiable or model-judged condition is met.

CA

Cursor Automations

App · Cursor (Anysphere)

2

Cursor feature that triggers cloud coding agents automatically on events such as a GitHub pull request being opened.

EL

ElevenAgents

App · ElevenLabs

1

ElevenLabs platform for building, deploying and monitoring real-time voice and chat AI agents.

CF

Coinbase for Agents

API · Coinbase

1

Coinbase's offering that lets AI agents trade, pay and hold funds through Coinbase accounts and APIs.

Claude Code mode that fans out many parallel sub-agents for deep tasks such as full codebase reviews.

AS

Agent Skills

Library

1

MIT-licensed pack of 25 engineering skills and slash commands for AI coding agents, covering spec through ship.

L3

Last 30 Days skill

Library

3

Open-source agent skill that searches Reddit, X, YouTube, HN, Polymarket and more for the last 30 days and summarizes it.

GN

Gemini Notebook

App · Google

5

Google's AI research and note-taking assistant (formerly NotebookLM) that answers from sources you upload.

ON

Open Notebook

App

1

Open-source, self-hostable alternative to Google NotebookLM with multi-provider AI, local models and multi-speaker podcasts.

ES

Email sending in ChatGPT

App · OpenAI

1

ChatGPT feature that drafts and sends emails directly from within a chat conversation.

OS

Oscar

Model

1

Action-conditioned robotics world model fine-tuned from Cosmos-Predict2.5-2B using 2D skeleton conditioning across embodiments.

MH

Midjourney hardware

Hardware · Midjourney

1

Unannounced first hardware product from Midjourney; no specifications or release details have been published.

AL

Agents Last Exam

Research

1

Benchmark of AI agents on long-horizon professional workflows across 55 sub-industries, led by Berkeley RDI.

MS

MiniMax Sparse Attention

Research · MiniMax

1

MiniMax's open-source sparse top-k and dense attention GPU kernels for NVIDIA SM100 (Blackwell), with a technical report.

VI

VideoMDM

Research

1

Diffusion framework that learns text-to-3D human motion generation from 2D poses extracted from monocular videos.

CO

ComfyUI-KJNodes

Library · kijai

2

Open-source ComfyUI custom node pack with utility, model optimization and quality-of-life nodes, including Set/Get.

HE

Headroom

Library

1

Open-source library, proxy and MCP server that compresses tool outputs, logs, RAG chunks and history before they reach an LLM.