AIYXL
中文SearchAIYXL StudioSearch

AIYXL MODELS

Models shaping
the next AI economy.

Track foundation models, multimodal systems, releases, capability shifts and the companies pushing the frontier.

63Model intelligence stories
14Tracked model families

FEATURED MODEL INTELLIGENCE

What matters now

Latest intelligence

Gemini

Latest news

Google's Gemini 3.8 Live Virtual Human Goes Live: Real-Time Lip Sync + Natural Expressions, Seamless Switching Across 97 Languages

On September 24, Google officially announced the launch of Gemini 3.8 Live with the Live Avatar feature, bringing near-real-time visual presentation to its native real-time conversation model. The feature combines low-latency video generation with speech to create an enterprise-grade interactive experience that can listen, observe, and communicate through a dynamic visual persona; it is now available on Gemini Enterprise.

MODEL FAMILIES

Follow the frontier

14 tracked families
01

GPT

234stories02

Gemini

207stories03

DeepSeek

198stories04

Claude

194stories05

Qwen

149stories06

Grok

60stories07

Kimi

56stories08

GLM

50stories09

Mistral

47stories10

MiniMax

42stories11

Llama

35stories12

Wan

15stories13

FLUX

13stories14

EXAONE

3stories

Family labels are an AIYXL editorial intelligence layer derived from historical coverage. Individual stories may reference more than one model family.

LATEST

Latest Model Intelligence

All intelligence

MODEL SIGNAL

Latest news

Amap Releases World's First 3D-Native City World Model ABot-Earth 0.7: 10-Minute Kilometer-Scale City Generation, 1,000x Efficiency Gain

On September 10, Amap (Gaode), a subsidiary of Alibaba Group, officially released ABot-Earth 0.7, the world's first 3D-native city world model. The model supports integrated full-scale AI generation from planetary view to street level, covering over 196 countries and regions—making it the most extensive digital Earth to date. Its official experience portal is now live, with capabilities already deployed in Flight Street View 2.0.

DeepSeek

Latest news

DeepSeek V4.1 Flash Officially Released: 552B Parameters Punching Above Its Weight, V4 Pro Fully Retired

9.10 On September 10, DeepSeek officially released the V4.1 Flash model, the smallest in its new architecture series, yet one that comprehensively surpasses its previous flagship V4 Pro. Alongside the launch, DeepSeek announced an orderly retirement of V4 Pro: after 12:00 on September 14, all V4 Pro requests will be routed to V4.1 Flash and billed at the new rates.

GLM

Latest news

Zhipu Open-Sources GLM-5.3-Flash: Ox Alpha Mystery Solved, Entirely Running on Domestic Chips at 1/40 the Price of Opus 4.8

8.27 Over the past week, an anonymous model named Ox Alpha on OpenRouter and OpenCode sparked intense global developer interest, setting new traffic records on both platforms. On August 26, the mystery was solved—Zhipu officially launched and open-sourced GLM-5.3-Flash (320B-A18B) , the first native multimodal model in the GLM-5 series. Even more striking, all traffic during the anonymous testing period was served entirely on domestic AI chips, with peak daily capacity reaching 100 trillion toke

Gemini

Latest news

Google Enters Legal AI Arena with Dedicated Gemini Tool, Taking on Thomson Reuters and Harvey

8.26 On August 25, Google Cloud officially launched Gemini Enterprise for Legal, an enterprise-grade, purpose-built agentic AI solution engineered for law firms and lawyers, marking Google's formal entry into the legal tech sector-2-7. The solution includes specialized skills for legal brief drafting, citation verification, contract lifecycle management, regulatory horizon scanning, and data subject access request (DSAR) fulfillment, along with secure connectors to legal software and data platfo

Wan

Latest news

Alibaba Cloud Wan3.0 Video Model Officially Launches: 30-Second One-Shot Generation, Documents to Video

8.24 On August 24, Alibaba Cloud announced the official launch of its video generation model Wan3.0. In just 18 days since its public beta on August 6, the model has continuously evolved in instruction following, cross-shot consistency, audio quality, and video editing capabilities. Enterprise users have described it as "stable, realistic, and textured," and it has already been integrated into production workflows across short dramas, advertising, cultural tourism, and music video industries.

MiniMax

Latest news

MiniMax Design Lets You Direct Videos with Natural Language—No Timeline, No Keyframes

8.20 MiniMax officially launched the multimodal creation Agent workspace MiniMax Design, marking a shift in AI video creation from "manipulating pixels" to "manipulating semantics". The product is built on H3, the flagship video model the company open-sourced just a month ago, aiming to translate model capabilities into commercial productivity.

GPT / Claude

Latest news

Alibaba Releases Qwen-UI-Agent: Enabling AI to Truly "Use" Every Screen, Surpassing GPT and Claude Across Mobile and Desktop

8.20 Alibaba officially released Qwen-UI-Agent, a real-world-centric foundation GUI agent model covering mobile devices, desktops, web browsers, and DeepSearch environments. The model matches or surpasses flagship models like GPT-5.6 Sol and Claude Opus 4.8 across multiple core benchmarks, transforming AI into a "digital executor" capable of understanding screens and operating software.

EXPLORE MODELS

Understand the model
ecosystem.

01

Releases

New model generations, launches and version updates.

02

Capabilities

Reasoning, multimodality, agents and frontier capabilities.

03

Benchmarks

Performance signals, evaluations and model comparisons.

04

Deployment

Cost, efficiency, open source and real-world adoption.