Yes: MiniMax has publicly linked Space Bunny to the upcoming MiniMax-M3.1-Flash. Its October 19 teaser names both, and its API documentation now lists that model. This identifies the announced product, although it does not establish that every earlier anonymous preview used identical model weights.
Space Bunny attracted attention before anyone put a company name behind it. The anonymous model appeared on OpenRouter in September and was reported at the top of its daily token-usage chart. MiniMax's October 19 announcement gives that story a much clearer answer.
What did MiniMax actually announce?
MiniMax's official X account posted a short video with a direct message:
"Welcome back. Space Bunny. MiniMax-M3.1-Flash is coming."
The wording explicitly connects Space Bunny with MiniMax-M3.1-Flash. It goes further than a visual hint or a community guess about the model's developer. The post still says the model is coming; it does not give a launch date or establish that every account can already access it.
Which ranking did Space Bunny top?
The platform was OpenRouter, and the claim refers to its historical daily token-usage ranking. A September 29 hands-on article recorded Space Bunny's earlier rise to first place on OpenRouter and OpenCode.
OpenRouter ranks usage by the tokens processed through its API. A high position shows how heavily a model was used during that period. It does not establish the best coding quality, and a September daily result should not be presented as today's ranking.
What do the official model pages confirm?
MiniMax's current documentation names MiniMax-M3.1-Flash and describes its capabilities. OpenRouter's Space Bunny page still uses the anonymous stealth/space-bunny-alpha identifier. These are separate listings; the announcement does not make their identifiers interchangeable.
Official listings checked October 19, 2026
| Detail | Space Bunny Alpha | MiniMax M3.1 Flash |
|---|---|---|
| Listed identifier | stealth/space-bunny-alpha | MiniMax-M3.1-Flash |
| Context window | 1,000,000 tokens | 1,000,000 tokens |
| Input | Text, images and video | Text, images and video |
| Reasoning | Adjustable effort | Five effort levels: low, medium, high, xhigh, max |
| Developer attribution | Still anonymous on OpenRouter's page | Named in MiniMax's official docs |
The overlapping specifications provide context. The official post supplies the public connection between the names. Neither establishes that the final release is unchanged from every version served during the anonymous preview.
What should users check before switching?
Use the model identifier published by the provider you are calling, then check access and pricing on that route. For MiniMax's documented M3.1 Flash API, thinking stays on; reduce the effort level if you need less reasoning time. Confirm the result on a small coding task before moving a longer workflow.
The teaser and documentation were reviewed for this article. No API request or comparative benchmark was run.
Looking for a coding subscription? Explore coding plans and check the models included in each plan.