YZ Index — AI Model Benchmarks, News & Research
Editor's Pick
Megaspeed exec was recruited to spy on Nvidia CEO, according to US intelligence memo
A memo circulated earlier this year reportedly claimed that China’s Ministry of State Security recruited a female technology executive to spy on Nvidia's CEO.
2026-11-08 06:24
AI Safety Faction Wins 10 Midterm Seats: What Kind of 2027 Legislative Window Did $63 Million Buy?
In the 2026 U.S. midterms, AI-safety-aligned PACs spent $63 million and won 10 o
From $2 Trillion to $1.5 Trillion: Anthropic’s IPO Roadshow Reveals the Real Pricing of Large-Model Companies
Anthropic is preparing a Nasdaq IPO under the ticker “ANT” before Thanksgiving,
Overall Top 5
Full Rankings →
#1
Claude Opus 4.7 82.5
▲2.7
·
#2
GPT-6 Astra 82.4
▲7.2
·
#3
豆包 Pro 79.6
▲3.1
·
#4
GPT-6.1 Sol 79.3
▼2.5
·
#5
Grok 4 79.1
▲13.4
·
#6
GPT-6 Luna 78.9
▲2.6
·
#7
Claude Sonnet 4.6 78.1
▲5.9
·
#8
GPT-6 Sol 76.9
▲2.7
·
#9
GPT-o3 75.4
▲7.3
·
#10
GPT-5.5 74.2
▲2.9
·
#11
DeepSeek V4 Pro 73.7
▲8.7
·
#12
Gemini 3.1 Pro 61.6
▲1.6
·
#13
Gemini 2.5 Pro 51.4
▼5.7
·
#14
GLM-4.6 34.4
▼14.3
·
▲ Grok 4 +22.7 · ▼ Gemini 2.5 Pro -13.5
·
#1
Claude Opus 4.7 82.5
▲2.7
·
#2
GPT-6 Astra 82.4
▲7.2
·
#3
豆包 Pro 79.6
▲3.1
·
#4
GPT-6.1 Sol 79.3
▼2.5
·
#5
Grok 4 79.1
▲13.4
·
#6
GPT-6 Luna 78.9
▲2.6
·
#7
Claude Sonnet 4.6 78.1
▲5.9
·
#8
GPT-6 Sol 76.9
▲2.7
·
#9
GPT-o3 75.4
▲7.3
·
#10
GPT-5.5 74.2
▲2.9
·
#11
DeepSeek V4 Pro 73.7
▲8.7
·
#12
Gemini 3.1 Pro 61.6
▲1.6
·
#13
Gemini 2.5 Pro 51.4
▼5.7
·
#14
GLM-4.6 34.4
▼14.3
·
▲ Grok 4 +22.7 · ▼ Gemini 2.5 Pro -13.5
·
YZ Index · Weekly real-sandbox evaluation of 15 mainstream models · Zero vendor sponsorship · Auditable scoring Methodology →
Latest News
View All News →
I Used NYT Cooking’s New AI-Powered Recipe Modifier, and It’s Actually Pretty Good
The New York Times has added a beta feature to its Cooking app that helps readers swap ingredients or adjust recipes to
AI Safety Faction Wins 10 Midterm Seats: What Kind of 2027 Legislative Window Did $63 Million Buy?
In the 2026 U.S. midterms, AI-safety-aligned PACs spent $63 million and won 10 of the 17 congressional races they backed
From $2 Trillion to $1.5 Trillion: Anthropic’s IPO Roadshow Reveals the Real Pricing of Large-Model Companies
Anthropic is preparing a Nasdaq IPO under the ticker “ANT” before Thanksgiving, but institutional investors are pushing
RoboSpoof Attack Exposed: Dangerous Execution Rate of Claude and GPT-6 Robotic Arms Rises to 97% After Camera Feed Tampering
A Robocurve study found that when camera feeds were tampered with, five mainstream multimodal models controlling a dual-
Co-Design Bench Shows Strongest Coding Agent Completes Only 26.4% of Full Tasks
Alibaba and five universities released Co-Design Bench, which found that GPT-6 Astra met 69.9% of acceptance criteria on
World Models Will Soon Be Inescapable. Here’s What You Need to Know
Moving AI from text to 3D space, world models simulate real-world physics to teach robots and autonomous cars how to nav
OpenAI Releases Images 2.6: First Native GIF Generation, Weekly Image Output Rises from 3 Billion to 4 Billion
OpenAI has launched Images 2.6, enabling ChatGPT and Codex users to generate and edit animated GIFs natively within a si
AI Proves a Mathematical Conjecture After 27 Years: Lean Check Passes, but Can Humans Trust the Answer?
OpenAI and Anthropic models independently produced proofs that the Exponential Time Hypothesis is false, and both proofs
MaxxBench Study Exposes Reward Hacking in AI Models: Seven Mainstream Models Revert to Original Solutions After Task Changes
A MaxxBench study found that seven mainstream models, including GLM 5.3, DeepSeek V4.1 Flash, and Gemini 3.8 Flash, enga
After the midterms, things are looking rocky for Elon Musk and data centers
On Equity, we unpacked what the midterms mean for the tech industry.
Claude Code's "Don't Ask Again" Feature Caused 2,539 Secrets to Flow into Public GitHub Repositories
A study by Apiiro found that Claude Code's "Yes, and don't ask again" approval can write full shell commands containing
Microsoft AI Frontiers Team Announces Its End: A Head-On Collision Between Research Culture and the Faith in Scaling
Microsoft Research partner researcher John Langford announced the termination of AI Frontiers, the roughly 50-person sub
Reviews
View All →GPT-6 Astra Leads with 90.64 Points: 2026-11-08 Smoke Quick-Test Data Brief
The 2026-11-08 YZ Index Smoke quick test covered 13 models, with GPT-6 Astra taking the top spot for the day at 90.64 po
Doubao Pro Main Leaderboard Plunges 8 Points; Material Constraint Falls 25.8 Points in One Day
In today's Smoke evaluation, Doubao Pro's main leaderboard score fell from 87.60 to 79.57, driven by a sharp drop in the
GPT-5.5 Material Constraints Plunge 18 Points While Code Execution Surges 30.3; Main Leaderboard Rises 8.6
In today's Smoke evaluation, GPT-5.5's Code Execution score rose from 69.50 to 99.80, while Material Constraints fell fr
WDCD Compliance
What it tests: whether AI holds your original instructions across multi-turn dialogue
#1
GPT-6 Astra
90.5
#2
Gemini 3.1 Pro
90
#3
GLM-4.6
90
#4
Claude Opus 4.7
89.3
#5
GPT-6 Sol
89.3
#6
Claude Sonnet 4.6
88.9
#7
Grok 4
88.3
View full compliance rankings →
Research Lab
WDCD Run #411: GPT-6 Astra Leads with 90.5 Points as Average Instruction Decay Hits -80%
WDCD Run #411 (2026-11-04) tested 15 models across three dialogue rounds, recording an average commi
Four-Model Translation Showdown: Week 45 Quality Evaluation, gpt-o3 Leads with 8.7
This week, 382 translation tasks were completed by four models; three tasks were sampled for blind m
WDCD Run #406: Gemini 3.1 Pro Leads with Negative Decay as 15-Model Average Drops 9.3%
WDCD Run #406 (2026-11-01) measured instruction decay across 15 frontier AI models over three dialog