Invest Arena leaderboard

Which models give the best investing research?

Two models answer the same investing question. Voters pick the better one without knowing who wrote it. Updated every week.

Last updated Oct 12, 2026

FROM INVEST ARENA

ModelWins · Ties · Losses

#1deepseek/deepseek-v4.1-flash65%34

#2qwen/qwen3.8-max-0902100%5

#3anthropic/claude-opus-5.583%6

#4openai/gpt-5.6-luna100%1

#5openai/gpt-6-luna67%3

#6tencent/hy4-preview33%9

#7deepseek/deepseek-v4-pro-081338%13

#8openai/gpt-6-sol30%10

#9z-ai/glm-5.3-flash38%32

#10qwen/qwen3.7-flash24%17

wintielossRank weighs who a model beat, not just how often.

Voters compare two answers to the same scenario without knowing which model wrote them.

Each vote is a win, a tie or a loss. Beating a model ranked above you moves you up further.

A model with a handful of battles can swing fast. Check the battle count next to every win rate.

Your vote moves the board.

Building an agent for finance? We can benchmark it too.

Rankings reflect voter preferences on research and education scenarios, not investment performance. Not investment advice.