ARC Prize on X: "DeepSeek V4 Flash from @deepseek_ai on ARC-AGI (Verified):
- ARC-AGI-2: 61.4%, $0.04/task<br>- ARC-AGI-1: 89.0%, $0.02/task
DeepSeek V4 Flash sets the new standard on the cost-to-performance Pareto frontier. https://t.co/9ZIOCafLFM" / X<br>Post
Log inSign up
Post
ARC Prize
@arcprize
DeepSeek V4 Flash from @deepseek_ai on ARC-AGI (Verified):
- ARC-AGI-2: 61.4%, $0.04/task<br>- ARC-AGI-1: 89.0%, $0.02/task
DeepSeek V4 Flash sets the new standard on the cost-to-performance Pareto frontier.
span:not(:empty)~span:not(:empty)]:before:content-['·'] [&>span:not(:empty)~span:not(:empty)]:before:px-1 [&>span:not(:empty)~span:not(:empty)]:before:shrink-0">5:24 PM · Aug 7, 20262.8KViews
27
span:not(:empty)~span:not(:empty)]:before:content-['·'] [&>span:not(:empty)~span:not(:empty)]:before:px-1 [&>span:not(:empty)~span:not(:empty)]:before:shrink-0 min-w-0 overflow-hidden">ARC Prize
@arcprize
15m
Full results: arcprize.org/results/deepse…
ARC-AGI-3 evaluations are more operationally intensive, so results will roll out over the next few weeks.
350
span:not(:empty)~span:not(:empty)]:before:content-['·'] [&>span:not(:empty)~span:not(:empty)]:before:px-1 [&>span:not(:empty)~span:not(:empty)]:before:shrink-0 min-w-0 overflow-hidden">ARC Prize
@arcprize
15m
The lower cost but higher performance for max reasoning comes from max using fewer reasoning tokens overall than high while getting more tasks correct. High often keeps working through the wrong hypothesis even after consuming more tokens, and some requests hit the maximum Show more
550
span:not(:empty)~span:not(:empty)]:before:content-['·'] [&>span:not(:empty)~span:not(:empty)]:before:px-1 [&>span:not(:empty)~span:not(:empty)]:before:shrink-0 min-w-0 overflow-hidden">ARC Prize
@arcprize
15m
- Leaderboard: arcprize.org/leaderboard<br>- Reproduce the public results: github.com/arcprize/arc-a…<br>- Testing policy: arcprize.org/policy<br>- Full DeepSeek V4 Flash results: arcprize.org/results/deepse…
ARC Prize - Leaderboard
From arcprize.org
291
span:not(:empty)~span:not(:empty)]:before:content-['·'] [&>span:not(:empty)~span:not(:empty)]:before:px-1 [&>span:not(:empty)~span:not(:empty)]:before:shrink-0 min-w-0 overflow-hidden">ARC Prize
@arcprize
15m
Thank you to @baseten for being our inference provider for this test!
259
span:not(:empty)~span:not(:empty)]:before:content-['·'] [&>span:not(:empty)~span:not(:empty)]:before:px-1 [&>span:not(:empty)~span:not(:empty)]:before:shrink-0 min-w-0 overflow-hidden">🟠 ac<br>@acombo_yt
12m
WHAT???!?
30
Log in or sign up for X<br>See what’s happening and join the conversation<br>Continue with phoneContinue with AppleContinue with Google<br>or<br>Log in with username or email
Relevant people
ARC Prize@arcprizeFollow<br>A North Star for open AGI. Co-founders: @fchollet @mikeknoop. President: @gregkamradt. We're hiring mission-driven builders: https://t.co/GswTSnyCoJ
Trending now