None
None
Long Story Video Skill
Turn your idea into 10s–10min video—AI writes scripts, prompts, and generates footage automatically.
Ads Video Skill
Generate professional ads and sales videos—AI auto-generates scripts, prompts, and footage.
3D Science Explainer Video Skill
Convert scientific concepts into stunning 3D explain animations
Feedback
freeTrialImage.bannerPity
freeTrialImage.upgradeUnlock
- ✓freeTrialImage.benefitHd
- ✓freeTrialImage.benefitWatermark
- ✓freeTrialImage.benefitUnlimited
AI Ad Video Example
Loading...
Space Bunny AI Model
Space Bunny AI Model opens the free space-bunny-alpha preview: a 1M-token context window, image and video prompts, and every spec traced to its source.
All Tools
Discover our comprehensive AI-powered animation toolkit
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.
Seedance 2.0
The Future of AI Video Is Here.
FLUX 3 Video Generator
Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
Seedance 2.0
The Future of AI Video Is Here.
A Closer Look at the Space Bunny AI Model Preview
The Space Bunny AI Model is a codename-only launch hosted on TokenRA: the lab behind it stays undisclosed, access costs nothing while the preview runs, and each figure below carries a note on where it came from.
- A Codename Listing With No Vendor AttachedThe Space Bunny AI Model is filed as space-bunny-alpha on TokenRA, where the provider stays anonymous and the codename is the only handle — there is no vendor page, model card or press release to check.
- Multimodal Input and a One-Million-Token MemoryThree details set the Space Bunny AI Model apart: it reads video as well as still images, it carries a 1,000,000-token context with output capped at 524,288 tokens, and it charges nothing for the duration of the preview.
- Every Figure Is Tagged With Its OriginSpeculation runs high around unreleased models, so this Space Bunny AI Model page separates what the provider states from what third parties measured and what remains pure guesswork — each line marked with its source.
Three Steps to Your First Space Bunny AI Model Call
Everything it takes to move from a blank TokenRA account to a working space-bunny-alpha call, in three moves.
What Space Bunny AI Model Can Do
A walk through the space-bunny-alpha spec sheet: context size, multimodal input, streaming error behaviour, reasoning controls and the zero-cost preview window.
One Million Tokens of Working Memory
The listing claims a 1,000,000-token context and a 524,288-token ceiling on output; as yet nobody outside the provider has published a replication of either number.
Image and Video Prompts Accepted
Text, stills and video all go in, and text comes back out — but vision is the thinnest-documented surface here, so treat it as a bonus rather than a settled capability of Space Bunny AI Model.
A Reasoning Effort Dial You Control
Reasoning effort governs both quality and latency: pushing it higher spends more thinking tokens before the answer appears, which shows up clearly against the 80 tok/s P50 recorded on the 24 September 2026 snapshot.
Errors That Surface Mid-Stream
Once streaming has begun a failure cannot be expressed as an HTTP status, so TokenRA emits it as a server-sent event carrying finish_reason: "error" — handlers that only inspect the opening status will cut output short without warning.
Zero-Cost Access During the Preview
Pricing shows $0 for the preview period, with no signup wall, model card or card-on-file requirement — though past free stealth windows have closed the moment a model was claimed.
Tokenizer Fingerprints Point Elsewhere
The clearest public clue is a measurement rather than an announcement: probes run by the stealthprint project match the MiniMax tokenizer family, which shrinks the list of possible labs without confirming one.
Space Bunny AI Model: Questions Answered
Straight answers on the space-bunny-alpha listing, what it can do, and how long the free window may stay open.
Does Space Bunny Alpha cost anything?
No — at the time of writing the listing shows $0 for the preview and asks for no payment details, though every free stealth window so far has ended once the model was claimed and repriced.
Which lab is behind Space Bunny Alpha?
Still unconfirmed. The provider field is blank, the tokenizer is tagged only as "Other" and the knowledge cutoff is empty, so the MiniMax M3.1 link remains a hypothesis built on tokenizer-family evidence, not a vendor statement.
Which inputs can I send?
Text, images and video are all accepted, with text returned — a combination that draws much of the interest, even though multimodal handling is the least documented part of the listing.
How large is the context and how much can it output?
The provider reports a 1,000,000-token context with a 524,288-token output cap; no independent reproduction of either figure has appeared yet.
What do I need before my first API call?
Just a TokenRA API key and an OpenAI-compatible client aimed at the base URL with the model slug — and remember that a negative balance blocks even free models until credits are topped up.
How long do free stealth windows tend to stay open?
Every stealth Alpha model tracked since early 2026 has eventually been claimed, mostly by Chinese labs, usually around a week after launch; Ox Alpha ran free for roughly six days. That pattern is context, not a forecast for this model.
Run Space Bunny AI Model Before the Free Window Closes
Open the live TokenRA listing, follow the step-by-step API guide, and build for the day the model is revealed and repriced instead of waiting for it — and never send data you would not hand to an anonymous third party.