AI Models & LLM News: Releases, Benchmarks, Comparisons
New models, benchmarks and our tests
Meituan's stealth model exposed: LongCat vs Qwen in 3 coding tasks
350M vs 685B: Liquid AI's tiny phone model takes on DeepSeek V3
Baidu unlimited-ocr meets Claude Opus 4.8: 1998 scan, clean table
We tested Sakana Fugu: 4 builds, zero retries, one prompt each
4 versions of GLM, same 3 prompts – each one beat the last
Liquid AI ships two tiny search models beating rivals twice their size
GLM 5.2 beats Opus 4.8 as Chinese labs sweep Hugging Face
4 Chinese models, 1 animation task: GLM spent 21 minutes and won
Rio 3.5: a Brazilian city hall just shipped a 397B open-weight model
Kimi-K2.7-Code vs MiniMax M3 on a frontend task