Skip to main content
MANIFOLD
When will Deepseek R2 be released?
53
Ṁ1kṀ30k
resolved Apr 27
Resolved
March-June 2026
100%98%
March-June 2026
0.2%
March-May 2025
0.2%
June-August 2025
0.2%
September-November 2025
0.2%
December 2025-February 2026
1.6%
July 2026 or later (never)

Does not have to be named R2 explicitly; would be whatever model that succeeds R1, likely model behind the deepseek-reasoner endpoint.

Market context
Get
Ṁ1,000
to start trading!

🏅 Top traders

#TraderTotal profit
1Ṁ1,740
2Ṁ349
3Ṁ234
4Ṁ178
5Ṁ159
Sort by:

@mods
It is March-June 2026

opened a Ṁ2,000 YES at 20% order

@Bayesian order up

@256 My credence is not lower than 20% unfortunately

oh i see now. No more coder after 2.5, no more reasoner after r1. They unify. Hmm

...and also I'm guessing 3.1 does not count too?

doesn’t R1 update count based on description? or nah? up to you

@NoAnswer is there any precedent for markets like these resolving for things like 4o or 2.5 pro updates? i have never seen that.

bought Ṁ8 YES

@256 i mean the description says this, which would imply that the deepseek r1 update counts. “Does not have to be named R2 explicitly; would be whatever model that succeeds R1, likely model behind the deepseek-reasoner endpoint.”

edit: lol i'm tired, totally thought this was my market

@NoAnswer a model successor is not the same as a model update

I guess based on market pricing, R1-0528 does not count.

@TotalVerb dont let that fool you im the only one pushing it down.

arb