MANIFOLD
Who will have the highest ranking model on web.lmarena.ai by EOY 2025?
17
Ṁ1kṀ1.8k
resolved Dec 23
ResolvedN/A
12%
OpenAI
57%
Anthropic
26%
Google
5%Other

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 2025-12-18 (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.

Market context
Get
Ṁ1,000
to start trading!
Sort by:

@Soli It seems this leaderboard will be sunset before the market closes

@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

@Bayesian makes even more sense, let's do this (FYI @traders )

@Soli That's a different dashboard with a different methodology though. The order is very different :P

@SqrtMinusOne hmmm good point, @Bayesian do you have a counter to this?

good point tbh. I checked the models that were in both leaderboards:

in common:

  1. opus-4-1

  2. sonnet-4-5-thinking

  3. gemini-2.5-pro

  4. minimax-m2

  5. glm-4.6

  6. sonnet-4-5

  7. qwen3-coder

  8. haiku-4-5

  9. 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)

bought Ṁ40 YES

@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

@Soli I'd N/A given that some people traded based on @Bayesian's proposal, which wasn't adopted but was added to the AI summary.

@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.

@prismatic > what's in it for you

Just fairness in the abstract :P

© Manifold Markets, Inc.TermsPrivacy