"Gemini 3" refers to a model recognized as the successor of Gemini 2 in a similar way to how Gemini 2 was the successor to Gemini 1, so Gemini 3.0 pro would count, but Gemini 2.X would not.
Models are considered released if they are available on an API endpoint or chat app option that does NOT include a term in the realm of "experimental", "preview", etc.
See also:
/Bayesian/when-will-google-release-gemini-3
/Bayesian/when-will-xai-release-grok-5
/Bayesian/when-will-openai-release-gpt6
/Bayesian/when-will-anthropic-release-claude-qZd5QEQ225
/Bayesian/when-will-deepseek-release-r2
/Bayesian/when-will-deepseek-release-v4
/Bayesian/when-will-meta-release-llama-5-6h9UypqOdp
/Bayesian/when-will-alibaba-release-qwen-4
/Bayesian/when-will-moonshot-release-kimi-k3
/Bayesian/when-will-google-release-nano-banan
/Bayesian/when-will-openai-release-their-next
/Bayesian/when-will-thinking-machines-announc
Update 2025-11-25 (PST) (AI summary of creator comment): The creator clarifies that the intent was to capture when "Gemini 3 Pro" as a class of models was considered released, because a "preview" is definitionally something not released.
The creator acknowledges there is ambiguity about whether Gemini 3 Pro is currently released or not, given the conflicting signals (chat app shows no "preview" label, but API and other sources do).
People are also trading
It’s not fair that you’re editing the description of a market that isn’t yours, especially when you hold a significant position in it.
With the updated description, it’s no longer clear that the author explicitly meant the exact release - just like in your other market about Gemini 3.0.
Yes, this market is basically just a duplicate.
The "Gemini Double Speak Version" refers to a text-to-speech (TTS) capability. It was introduced with the Gemini 2.5 models. This feature generates dialogue using two distinct voices.
This feature is also known as multi-speaker dialogue generation or multi-speaker text-to-speech. It allows users to create natural-sounding conversations between two different personas from a single text input.
Key details:
Models: It is available using the
gemini-2.5-flash-ttsandgemini-2.5-pro-ttsmodels.
@AffineTyped
I’m really asking you to please try to understand the question. The situation feels very confusing and frustrating. As we can all see, Gemini 3.0 has already been released, and the similar market mentioned in the rules was resolved YES a long time ago.
So it’s hard to understand why our funds here still need to remain frozen instead of being available for us to use elsewhere. Could you please unfreeze the funds and resolve this market appropriately?
For context, Gemini 3.0 was officially released publicly on 18 November, and it began rolling out in the Gemini app globally within the following three days.
Thank you for taking the time to resolve this
People are genuinely hesitant to buy here because of the confusing phrase “Gemini Double Speak Version” in the heading. Don’t understand what it is supposed to mean, and it’s creating uncertainty.
From the Gemini team’s response, it seems clear that the question you asked was about whether the next version would be 2.9-tts, not a completely new major release like Gemini 3.0.
But in reality, Google released an entirely new series: Gemini 3.0.0.
We’re just hoping this can be clarified so people feel safe interacting with the market. Thanks for taking the time to look into it.
Another example of Google DeepMind calling Gemini 3 released: https://youtube.com/shorts/KzB0ywf4V-E?si=f69vX3mpgSThfyKi
Gemini 3 (Google) release (No Double Speak Version) before 2025-11-21?
Yes: Gemini 3 (by Google) was released before November 21, 2025. The public launch was on November 18, 2025.
"The creator clarifies that the intent was to capture when "Gemini 3 Pro" as a class of models was considered released, because a "preview" is definitionally something not released."
Has no free tier, severely nerfed Tier 1 RPM (3x lower) and RPD (10x lower) rate limits. It’s about 3x slower to complete tasks relative to 2.5 pro on a wall clock basis due to it not being fully rolled out yet.


@Paul They use multiple name "3 Pro", "Gemini 3 Pro", and "Gemini 3 Pro Preview", "gemini-3-pro-preview" for the same model across multiple services (chat app, API..). And the status for this model is "Preview".
Source:

Exact same pattern for "Nano Banana 3 Pro Preview".
@Paul it called "thinking" , and slow by design.
That was a model that was fully released to the public on 18 November after a month of beta testing.
There’s an even slower version called “deep thinking.”.
More expensive and 3 times slower :deep thinking: for very valuable Is available too
@notbayesian Mormally devs ar resting finding bugs and update models befire final releases.. but with gemini 3.0 pro even after about a month of testing, no issues were found that required . No approach to updating the model, so it was released as was in beta tests and it turned out to be solid right from the start.



