AI Models & LLM News: Releases, Benchmarks, Comparisons
New models, benchmarks and our tests
Anthropic's HLE lead hits 16 points – and it's still accelerating
FrontierCode: first bench to test if AI code is actually mergeable
Two open agent models in one day: one codes, one lives
Gemma 4 E2B QAT: Google squeezes an AI model to 0.84 GB
Liquid AI just dropped a tiny model that reads images and pulls out exactly the data you need
MiniMax M3 vs Kimi K2.6: when a 1-point gap is the whole story
Google launches Gemma 4 12B – nearly matches the 26B model on benchmarks, sometimes beats it, at less than half the memory footprint
7 of 16 top AI labs shipped a model in May 2026
Qwen drops 3.7 Plus – the model that reproduces apps on its own
5 releases, +0.3%: open-weight coding hit a wall