Artificial Analysis v4.3.2 to v5, AI models re-ranking step by step.

GPT has funded AA.

everytime openai falls behind aa changes their method

Terra above Kimi K3.1 and Fable 5.1? Bullshit

I think OpenAI paid them a money, so their model will sit higher and other open model sit lower.

Open-weight models showed the biggest changes. I guess they were overrated...

kimi was 4th like a week ago. Just brutal watching it slide

So everybody except from OpenAI gets nuked. 🤦 It doesn't touch Anthropic much, because people already know they're better. Open source gets degraded. When they're doing just the benchmark, you really have to manage any perception of favouritism better, fair or not, right??

Jew test looool

Chinese got cooked except kimi

Second time a new benchmark version pushes openai up. Nearly every LLM ranks lower

Vibe Bench

I dont agree with this at all

It seems a cluster for USA and China models, preferring OpenAi

So, am I to understand that this is a ranking based on your preferences? Haiku 5.5, the Muse and GPT 6.1 Terra models ranking higher than DeepSeek v4.1 Flash 1017 is a crime...

Best model is now worse than Haiku

Tbh, I knew Muse shouldn't have been that high 🥲🤣

This feels like a good direction towards deranking benchmaxed models without any deeper capability. The ranking changes do make sense from my testing. Gemini and Muse being in the top was some kind of unfunny joke.

Cool graphic bro

Только gpt выиграл

So now we know who's benchmaxxing and who isn't. Nice.👍

Do they just keep changing it for the better of OpenAI or do the chinese labs just benchmax?

Does the second tier actually move between versions, or just the top two?

Kimi dropped to tier 1.5 apparently and Grok/DeepSeek are now lead the second tier pack.

muse wonder above deepseek

V5 did absolutely nothing?

claude opus 5.5 "and still i rise".. meanwhile openai improving their models after each iteration

It's only an accurate benchmark when Argon is 20th.

We're iterating and improving every day, and I don't know which one to choose.

I agree to some extent.

muse who?

just a wild chicken index

Grok 😭

tbh, now it makes more sense

Fable 5.5 coming soon