The stealth/space-bunny-alpha model ID was withdrawn on October 5, 2026 and requests to it will fail. Original examples below are historical records, not runnable current integrations.
SPACE BUNNY / DEVELOPER GUIDE
Space Bunny Alpha API: start with one request
Developers previously accessed Alpha through OpenRouter's Chat Completions API. The retired route is archived here; current chat uses DeepSeek V4.1 Flash through the same API. This website is an independent chat interface; it does not sell or distribute model API keys. Use the workspace for current DeepSeek chat without writing an integration.
Prepare an account and key
Create an OpenRouter account and generate your own key. Store OPENROUTER_API_KEY in a server-side environment variable. Do not put it in a public repository, browser code or a URL. Read the model listing and Stealth terms before use, especially if you plan to serve other users.
Model IDs and where to use it
The archived request used OpenRouter model ID stealth/space-bunny-alpha. OpenCode Zen documentation separately listed a time-limited Space Bunny Free with ID space-bunny-free. These IDs belong to different platforms; do not interchange them or assume identical limits and data terms. If you do not need your own API key, sign in to use this site's chat.
Historical minimal text request
Send a POST request to the endpoint below with Content-Type: application/json and an Authorization: Bearer header containing your server-side key. This archived body will now return a model-unavailable error.
https://openrouter.ai/api/v1/chat/completions{
"model": "stealth/space-bunny-alpha",
"messages": [
{
"role": "user",
"content": "Explain a hash table with an example."
}
],
"reasoning": {
"effort": "low"
},
"max_tokens": 2048
}Add capabilities when you need them
- Set reasoning.effort to adjust depth. Higher settings can increase latency and consume more of the output budget.
- Enable stream for SSE responses, handling chunks, error events and the completion marker.
- Alpha previously accepted image_url and video_url content parts. For current DeepSeek use image_url only; native video is not listed.
- Use response_format for JSON. For tool calling, your application must execute requested tools and send their results back.
Historical terminal request (retired ID)
This command uses Bash syntax for macOS or Linux. Set OPENROUTER_API_KEY in your current terminal first; keep your real key out of saved examples.
curl https://openrouter.ai/api/v1/chat/completions \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"stealth/space-bunny-alpha","messages":[{"role":"user","content":"Reply with exactly HELLO."}],"max_tokens":256}'For a completed text response, read choices[0].message.content. Check the HTTP status and any error payload before reading choices. With tools enabled, a message may contain tool_calls instead of a final text answer.
What to check when a request fails
- 401: check that the server-side key is present and belongs to the intended OpenRouter account. Never log the key.
- 402: check account credits, billing details and whether the model is available to your billing region. This is not a JSON-formatting or prompt problem.
- 429 or timeout: inspect rate-limit details and retry with bounded backoff. A stream can also fail after it has started, so handle error events.
Use OpenRouter's response body and documentation for the specific cause. This site does not supply API allowance to third-party integrations.
Free preview is not an availability guarantee
Alpha listed zero input and output token prices during its preview, which ended October 5. Current DeepSeek API requests are paid and subject to account limits; this website separately funds a 50-reply daily free allowance. Handle 429 responses, timeouts, empty answers and interrupted streams without unbounded retries. Avoid sensitive material because the provider may retain inputs and outputs. Confirm the terms permit your intended public service.
Which model ID can replace the retired route?
This site now uses deepseek/deepseek-v4.1-flash through OpenRouter. This is an independent replacement, not Alpha's revealed identity or official successor. Keep the endpoint and server-side authentication, replace model, and test a short text request before images, tools or JSON. DeepSeek API access is paid; do not copy video_url into the new request.
{
"model": "deepseek/deepseek-v4.1-flash",
"messages": [
{
"role": "user",
"content": "Reply with exactly HELLO."
}
],
"max_tokens": 256
}DeepSeek V4.1 Flash specifications ↗404 / No endpoints found for the retired ID is not a prompt problem. The 401, 402 and 429 troubleshooting below remains applicable to the replacement.