Finally updated PoliBench with Fable 5.1 GPT-6.1 Sol, Claude Opus 5.5, Gemini 3.8 Flash & Grok 4.7. Plus a ton more. See where the models land. polibench.jonathanrreed.com/…

GPT 6.1 Terra seems like a great advertisement for Sol 6.1 Medium

The internet really feels like it’s been breaking at the seams recently

Anyone else seeing this in chatgpt?

Grok 4.7 is pretty different from the other recent grok models, while Anthropic has been pretty steady.

I feel like the Gemini 4 hype lasted less than a minute

is 6.1 Sol coming to chat? Its still stuck on 5.6 Sol

Seeing Sonnet above Astra feels weird, and wild too after the pricing changes.

Feel bad halving the 200 tier of codex the same day they announced the 500 their plan.

Tibo unjammed the button! The reset just came in!!

Did the button get jammed? @thsottiaux

I wonder if the new ChatGPT 24/7 bot will be on the same limits as Codex/Work. With the rumors that ChatGPT limits are merging with Codex/Work limits, it could make even the potential $500 plan feel restricted

Crazy day seeing SimpleBench from @AIExplainedYT become pretty much saturated. Models are almost at the highest human score

Luna 6 is going to be the sleeper hit of the day, while Opus 5.5 is going to be what OpenAI wanted Sol 6 to be

I think Cloudflare is the only one to come out of the Union Alpha model chaos in a positive light, and it was probably by accident

You can really tell an area’s wealth by how many trees they have

I'm token broken ):

Cute lil snake game in the new chatgpt image gen if you click the animation