LATEST_
All posts in reverse chronological order
[4159] September 2026
Benchmarks
GPT-6 Astra Benchmarks: Three Tests Where It Loses
Benchmarks
GPT-6 Astra Test: Non-English Writing, Code, 150K Context
Benchmarks
Imagen 3 vs Imagen 3 Fast vs DALL-E 3: 6-Task Test
Benchmarks
GPT-6 Astra Hallucination Test: Fake Facts vs Real Ones
Benchmarks
Claude vs GPT-6: Polish Text Costs 56% More Tokens
Benchmarks
GPT-6 Astra for Coding: What the Benchmarks Show (and Hide)
Benchmarks
ChatGPT Images 2.5 vs 2.0: What Actually Changed
Benchmarks
ChatGPT Image Editing Test: 11 Edits, 11 Unwanted Changes
Benchmarks
Qwen3.8 Max Test: 13 Tasks, Scores, Pricing & Access
News
Bank of England: We Must Retain Right to Intervene in AI
AI Tools
AI Logo Generator Test: Six Variants, One Hidden Problem
Benchmarks
GLM 5.3 Real Cost: Three Numbers Don’t Match the Model Card
Guides
OpenRouter Pricing: The Real Cost Can Be 14x Higher
Benchmarks
I Built an AI Test for 11 Models. One Task Had a Bug
Benchmarks
How Much Does an AI-Generated Article Cost? I Paid and Counted
Benchmarks
AI Context Window Test: Two Models Failed at 25,000 Tokens
Benchmarks
New AI Models Force Reasoning On - 7 of 18 Won’t Let You Disable
Guides
AI Photo Editing: Change One Thing - 3 Tests
Benchmarks
I Asked 6 AI Models 8 Biology Questions. Zero Refusals
Benchmarks
LLM Confidence Scales: 0-20 Works Better Than 0-100
Benchmarks
PIDS-Bench: Why Prompt Injection Detectors Fail in Real Use
Benchmarks
AI Agent Consistency: IBM Research Tackles the Reliability Gap
News
Chatbot Hallucination Nearly Triggered US Military Strike
Guides