Reasoning you can read
Space Bunny Alpha always thinks before it answers. Space Bunny streams that reasoning in a collapsible panel, then the final answer.
Space Bunny Alpha · stealth model chat
Space Bunny Alpha is the anonymous stealth model with a 1M-token context, strong coding and reasoning you can see. Space Bunny gives it a clean chat in your browser: pick how hard it thinks, watch the reasoning stream, keep every conversation.
THE MODEL
Space Bunny Alpha is a stealth model on OpenRouter: a large model released under a code name so developers can test it before its maker is announced. It is described as fast, strong at coding, natively multimodal and able to reason at an adjustable effort level, with a 1M-token context window.
Reasoning is mandatory, so every Space Bunny Alpha answer starts with a thinking pass. On Space Bunny you choose Fast, Balanced or Deep to trade speed for depth. The web chat sends text; the Space Bunny Alpha API also accepts images.
Stealth models can change or be withdrawn without notice, and their providers may log prompts to improve the model. Do not share secrets or personal data in Space Bunny chats.
Source: OpenRouter model listing, checked September 28, 2026.
WHY SPACE BUNNY
Space Bunny Alpha always thinks before it answers. Space Bunny streams that reasoning in a collapsible panel, then the final answer.
Fast for quick answers, Balanced by default, Deep for hard bugs and proofs. Every depth costs the same single credit.
Paste a function, an error or a spec. Replies render Markdown with fenced code blocks you can copy.
Chats are saved to your account. Reload or come back later and pick up the same thread.
HOW IT WORKS
Ask a question, paste code or start from an example prompt.
Fast, Balanced or Deep. Deep gives Space Bunny Alpha more room to think.
Reasoning streams first, then the answer, formatted in Markdown.
Follow up in the same chat; earlier turns are sent as context.
FOR DEVELOPERS
Use Space Bunny Alpha in your own app through an OpenAI-compatible endpoint. Keep your SDK, change the base URL and model, and pay 1 credit per request from the same balance as your chat. Streaming, reasoning effort, tool calls, JSON output and image input are supported.
from openai import OpenAI
client = OpenAI(
base_url="https://spacebunny.pro/api/v1",
api_key="sk_spacebunny_…",
)
reply = client.chat.completions.create(
model="space-bunny-alpha",
reasoning_effort="medium",
messages=[{"role": "user", "content": "Hi!"}],
)AT A GLANCE
Space Bunny is built for trying Space Bunny Alpha on real work: code, math, plans and long documents, with the reasoning visible.
An anonymous stealth model listed on OpenRouter. Its developer has not been announced.
Space Bunny sends your recent turns within a fixed per-message budget.
Reasoning depth does not change the price. Failed replies are refunded.
POST /api/v1/chat/completions with streaming, tools, JSON output and image input. 1 credit per request.
FAQ
Space Bunny (spacebunny.pro) is an independent web chat where you can talk to Space Bunny Alpha in your browser, choose its reasoning depth and keep your conversations.
Space Bunny Alpha is an anonymous “stealth” large model listed on OpenRouter. It is described as fast, strong at coding, with adjustable reasoning effort and a 1M-token context window.
The developer has not been announced. Stealth models are released under a code name for testing. Space Bunny is not affiliated with the model developer or with OpenRouter.
New accounts get 20 welcome credits. Each reply uses 1 credit. See Pricing for available credit plans.
Yes. Reasoning is always on for this model. Space Bunny shows the reasoning text it streams in a collapsible “Thinking” panel above each answer.
Chats are stored in your account and you can delete them. Messages are sent to OpenRouter and the stealth model provider to generate replies; stealth providers may log prompts to improve the model, so do not share sensitive or confidential information.
Yes. Create a key on the Space Bunny Alpha API page and call https://spacebunny.pro/api/v1/chat/completions with any OpenAI-compatible SDK, using the model space-bunny-alpha. Each request uses 1 credit from the same balance as the chat; failed requests are refunded.
The API accepts images (https URLs or base64 data URLs). The web chat currently sends text only. Video input is not offered yet.
New accounts get 20 free replies. No card needed.
Open the chatSpace Bunny is an independent service and is not affiliated with OpenRouter or the developer of Space Bunny Alpha. Model replies can be wrong; check important facts and code.