
LMSYS released a new leaderboard for ranking ai-models based on their web dev skills. It is super fun to play around with. Give it a shot here.
Update 2025-12-04 (PST) (AI summary of creator comment): The market may be resolved approximately 18 days early (around mid-December 2025) due to the leaderboard being sunset before the original close date.
Update 2025-12-05 (PST) (AI summary of creator comment): The market will resolve based on https://lmarena.ai/leaderboard/webdev instead of the original web.lmarena.ai URL, as the original leaderboard is being sunset before the market close date.
Update 2025-12-18 (PST) (AI summary of creator comment): The creator is reconsidering the resolution criteria and is currently leaning towards one of the following options:
Resolving based on the last status of the old dashboard.
Resolving the market as N/A.
People are also trading
@SqrtMinusOne i think it is fine, we are almost at the end of the year anyways. @traders any objections to resolving the market ~18 days earlier?
@Soli i would just think this should clearly resolve to https://lmarena.ai/leaderboard/webdev which is clearly meant to be where that leaderboard is moved for consistency around their leaderboard location conventions
@Soli That's a different dashboard with a different methodology though. The order is very different :P
good point tbh. I checked the models that were in both leaderboards:
in common:
opus-4-1
sonnet-4-5-thinking
gemini-2.5-pro
minimax-m2
glm-4.6
sonnet-4-5
qwen3-coder
haiku-4-5
grok-code-fast-1
Order in the old version
123456789
Order in the new version
216547839
one is a good proxy of the other, but not perfect and sometimes pretty bad (gemini 2.5 pro)
@Soli Before I buy in more, I wanted to double check my assumption that the dashboard Bayesian linked is the one this market resolves with, since the other one is now unavailable?
@prismatic not sure at this point but leaning towards either (1) resolvong according to last status of old dashboard orr (2) naing the market
@Soli could we get a resolution since market is closed? I would prefer option 1 since that's generally what other markets with leaderboards that have stopped being update do
@SqrtMinusOne I think the AI summary is pretty clear (Anthropic is probably going to win in either old dashboard or waiting EOY on new dashboard since there aren't any major model releases rumored for late December). Also regardless, traders trading on the AI summary instead of the comments is their own decision that shouldn't affect a resolution.
@prismatic > Anthropic is probably going to win in either old dashboard
gpt5.1 was the top-1 in the old dashboard as I recall. I had bought OpenAI on December 5 based on that, which I later sold.
@SqrtMinusOne oh ic if you sold what's in it for you? an N/A on the market isn't going to reverse that.
