What will be true about GPT-5?
296
6.6kṀ74k
2026
17%
OpenAI will claim it faster than GPT-4 Turbo
77%
It will have a different logo color from green, black, or purple(based on resolution of https://manifold.markets/MiraBot/what-color-will-the-next-openai-llm?r=TWlyYUJvdA)
4%
Will be claimed to be AGI by the New York times up to 3 months after release.
1%
Will be claimed to be AGI by OpenAI up to 3 months after release.
1%
Will be claimed to be AGI by Wikipedia up to 3 months after release.
28%
It will use a new architecture meaningfully different from GPT-4
1.9%
There will be credible reporting that it is or was "excessively horny" either before or up to three months after release
13%
It will be ranked the highest model on the LMSys Chatbot Arena, and not overtaken by another model, 3 months after the release date.
4%
It will be able to pass jim's "agents benchmark"

"GPT-5" refers to a model named or presented as "GPT-5"(or a variation). This can be the model id, or a description like "the fifth generation of the GPT series". Benchmarks and features are not considered: Only how the model is presented by OpenAI.

Add your own options. One who creates an option is responsible for clarifying follow-up questions. @Mira may interpret them if not, or resolve the option NA.

If GPT-5 is not released by end of 2025 or insufficient information is presented, options resolve NO by default unless specified to resolve differently.

If GPT-5 is released, moderators should close this market for trading immediately.

Resolution tags. Place at the beginning of an option to change how it resolves.

  • Default: Option resolves NO by default at end of 2025

  • Ambiguous(2025): Option resolves 50% by default at end of 2025.

See also: /MiraBot/what-will-be-true-about-gpt45

Get
Ṁ1,000
to start trading!
Sort by:

@mods Can you guys resolve this since Mira deleted her account

it appears that the different GPT-5 models have various different knowledge cutoffs. The ChatGPT versions (thinking and non-thinking) report “June 2024”. But the API documentation says “May 30th, 2024”, “September 29, 2024”, or “September 30th, 2024” depending on the model version.

https://platform.openai.com/docs/models/compare

sold Ṁ91 YES

@BionicD0LPH1N

I'm not sure how this should resolve. When using it via ChatGPT, it's a model router, but when you're using it via the API, you're picking a model directly.

@SimonWestlake This reads to me like the models themselves (what you can access through the API) don't have a built-in router, so it's not a core part of GPT-5 itself. The gpt-5-chat model in ChatGPT UI is a separate model that analyzes the query and makes a route call to the appropriate version of GPT-5, but this wrapper model itself is not part of the release and cannot be queried directly (since it's a pass-through layer). So GPT-5 is packaged with a router, but it's not a router itself, and the router is likely an extension of their input safety layer which had always been a separate model.

@moozooh Yeah I think that's the most reasonable interpretation.

@SimonWestlake I am of the opinion that it should resolve YES. This is roughly what I had in mind when I created this option.

sold Ṁ5 YES

@VitorBosshard Disregard my original comment, I think this should resolve NO

@Mira How will questions like this resove if GPT-5 has configuration options that allow it to be on either side of the line?

@jim resolves YES if a one-sentence, plain-English prompt is found that allows GPT-5 to solve jim's agents benchmark at least 20% of the time.

@jim This is a wild benchmark. I have some level of general intelligence and there is no way I could do this.

@MalachiteEagle I don't expect "open"AI to disclose such information to be honest. They don't even publish parameter couts anymore let alone training data details.

@ProjectVictory maybe. Guess we'll find out. Think they like to sell a narrative around the models they release, that narrative might include something like this.

@mods And this one resolves NO.

@mods Resolves YES.

@mods Resolves NO, along with the option for May - October.

The new "GPT-4o" counts as GPT-4.5 in the paired market, since it was based on the name. "A model with separate branding in the GPT-4 series". So I've resolved the 4.5 option here to match.

"GPT-5" designation here is also based on the name. I will not be looking at benchmarks, only how the model is presented by OpenAI.

bought Ṁ50 NO

@Mira wait if you consider gpt4o as gpt4.5, will the model Jimmy calles gpt4.5 here resolve as gpt5? https://x.com/apples_jimmy/status/1833595024543781088

@Phill It has to be named GPT-5. Or presented as the "fifth generation of the GPT series". So that name is unlikely to count.

FYI: If gpt2-chatbot was GPT-5 then it will have already been "released". That seems unlikely, but it's important to note. Ditto for GPT-4.5.

@Mira this conflicts with your earlier clarification:

has to be described as GPT-5. Otherwise it's just an internal name

Resolve the April answer please

@chrisjbillington It may have been "released" and may soon be known as GPT-5, but it wasn't "released as GPT-5". It need only be "released" not "released and known to be released", so it can't be resolved yet.

We had a similar situation with the Mistral-medium leak: /Mira_/when-will-mistralmedium-weights-be It may have happened in January, but we don't know.

This does mean all of the release options are blocked until end of 2025, waiting for a statement on what this model is. Even if GPT-5 is released in a couple months, we can't resolve those YES because it might retroactively resolve an earlier "release" YES if later known. (but if nothing is learned, at end of 2025 it would resolve to the official release since earliers would resolve NO in absence of information)

@Mira that's total BS Mira

@Mira have you come around about this yet? If so, resolution please.

© Manifold Markets, Inc.TermsPrivacy