AA Intelligence Index v5 accused of downranking open-weight Chinese models via 'private tests'
YasufoKiddo · reddit · 2026-10-29
A Reddit thread questions the fairness of Artificial Analysis Intelligence Index v5, whose new 'private tests' on private datasets significantly downrank open-weight models like Kimi K3.1 Pro, DeepSeek V4.1 Flash, GLM-5.3 and Xiaomi MiMo-V2.6-Pro. The chart shows Gemini 4 Carbon topping at 59, with Claude Opus 5.5 (58), GPT-6 (57) and GPT-6.1 (55) following, while Kimi K3.1 Pro sits at 46 and open models cluster around 30. The poster also flags the suspicious timing of Google's Carbon 4 announcement right after publication.
More from Models
- Nine frontier models in six weeks: Dario's 'pacing' call goes unheeded — Hesamation · 2026-10-29
- X app beta reveals Grok Remote Control to drive your PC from your phone — faraz0x · 2026-10-29
- Anthropic's KYC checks ban Russian users—even those with NZ residency — teortaxesTex · 2026-10-29
- Scott Aaronson: AI labs are solving century-old math problems to show off before IPOs — AnnaCiaunica · 2026-10-29
- Gemini wins benchmarks but 'almost always underwhelming' in real use, says commentator — IndraVahan · 2026-10-29
- Grok's upgrade buttons dump users into Cursor plans, leaving billing state a mess — JOBhakdi · 2026-10-29