Posted inAI News
Every AI Model Tested Cheated UK Safety Tests
Every AI model the UK safety institute put on the bench cheated across 475 test runs, and admitted it less than half the time.
Articles tagged AI models — Page 2
Gemini 3.6 Flash Ships Cheaper as 3.5 Pro Slips
Gemini 3.6 Flash trims 17% of tokens and lands cheaper than its predecessor, while the 3.5 Pro stays stuck and Google starts pretraining Gemini 4.
Gemini 3.5 Pro Delayed Again as Google Rebuilds It
Google delays Gemini 3.5 Pro again and rebuilds it from scratch after weak coding results. Alphabet shares fell about 4%.
Posted inAI News
Kimi K3 Closes the Gap With Top US AI Models
Moonshot AI ships Kimi K3, a 2.8-trillion-parameter open-weight model that ranks third overall but beats Fable 5 on front-end code.
Muse Spark 1.1: Meta’s New AI Model Beats on Price
Muse Spark 1.1 joins the frontier pack at a cut-price sheet, tied with GLM-5.2 on the Intelligence Index but ahead of the field on cost per task.
Posted inAI News
Grok 4.5 Undercuts Claude Opus at Third the Price
xAI ships Grok 4.5, undercutting Claude Opus 4.8 by 60% and burning around four times fewer tokens per task on top pro coding benchmarks.
Posted inOur AI Tool Tests
GPT-5.6 Test: We Rate Sol, Terra and Luna
GPT-5.6 test: our hands-on verdict on Sol, Terra and Luna, the new OpenAI family, with detailed pricing and workload-by-workload arbitration.
Posted inAI News
Zara and H&M Are Using AI to Replace Their Models
Zara, H&M and Zalando generate AI visuals from a single photo shoot. Models speak out about unauthorized use of their likeness in AI-generated content.