Space Bunny Free: OpenCode Zen Stealth Model

Updated: 2026-09-30

Everything you need to decide: what Space Bunny Free does, who it fits, and whether it is free.

Space Bunny Free is a free stealth model from OpenCode Zen. You can inspect its published parameters and use its output directly; the provider does not disclose the underlying model.

This is a stealth model. The community has speculated about its origin, but the provider has not confirmed it. This page records only the configuration Pi and OpenCode publish.

Model ID: space-bunny-free
Provider: OpenCode Zen
API: openai-completions
Base URL: https://opencode.ai/zen/v1
Context: 1,048,576 tokens
Max output: 524,288 tokens
Input: text, image
Reasoning: ✓
Pricing: USD 0 input · USD 0 output · USD 0 cache read · USD 0 cache write

Free right now. Input, output, cache read, and cache write are all USD 0. This is a limited-time offer, so check the latest pricing before you commit.

The 1,048,576-token context window can hold large codebases, long session histories, several specifications, and extensive retrieved material.

The 524,288-token output limit supports long implementations, file-by-file migration plans, complete reviews, and multi-stage reasoning results.

Reasoning is enabled, making the model suitable for architecture trade-offs, difficult bugs, cross-module refactors, and multi-step implementation planning.

It accepts text and images. Paste code and requirements, or add screenshots, mockups, and diagrams as context.

low, medium, high, xhigh, and max pass through unchanged; off and minimal are unavailable in this configuration.

How to use it

1.Select or configure Space Bunny Free in OpenCode using the model ID space-bunny-free and your OpenCode Zen API key.
2.Pi lists the API as openai-completions and the Base URL as https://opencode.ai/zen/v1; use OpenCode’s official installer and model-selection flow when possible.
3.Add the task, constraints, code, and relevant images to the prompt, then choose an appropriate reasoning effort.
4.Review the generated code, image-based conclusions, and reasoning steps; validate the result with tests or your real workflow.

Use cases

Cross-file reading, explanation, and refactoring in large codebases—the 1M window leaves room for more supporting context.
Analyze UI screenshots, architecture diagrams, or error screenshots alongside the relevant code.
Run long research sessions, compare documents, and iterate across many turns without frequently discarding earlier context.
Draft migration plans, long reviews, test strategies, or complete implementations.
Prototype, run batch evaluations, and trial workflows when budget matters—the current USD 0 pricing makes it easy to evaluate value.

FAQ

What is Space Bunny Free?

A free stealth model from OpenCode Zen with the model ID space-bunny-free. The provider publishes parameters and capabilities, but not the underlying model.

Is it free now?

Yes. Pi currently lists USD 0 for input, output, cache read, and cache write. It is a limited-time offer, so check the latest pricing before you commit.

Does it accept image input?

Yes. The official configuration lists text and image input, so you can add screenshots or other visual material as context.

Why are off and minimal reasoning levels unavailable?

The official thinkingLevelMap maps off and minimal to null, while low through max pass through unchanged. The usable levels are low, medium, high, xhigh, and max.

Is it MiniMax M3?

Its origin has not been officially confirmed. The community has speculated, but this page does not present an unverified identification as fact.

Configuration

{
  "providers": {
    "opencode": {
      "apiKey": "YOUR_API_KEY",
      "models": [
        {
          "id": "space-bunny-free",
          "name": "Space Bunny Free",
          "reasoning": true,
          "input": [
            "text",
            "image"
          ],
          "thinkingLevelMap": {
            "off": null,
            "minimal": null,
            "low": "low",
            "medium": "medium",
            "high": "high",
            "xhigh": "xhigh",
            "max": "max"
          },
          "contextWindow": 1048576,
          "maxTokens": 524288,
          "cost": {
            "input": 0,
            "output": 0,
            "cacheRead": 0,
            "cacheWrite": 0
          },
          "compat": {
            "supportsStore": false,
            "supportsDeveloperRole": false,
            "supportsStrictMode": true,
            "maxTokensField": "max_tokens"
          }
        }
      ],
      "api": "openai-completions",
      "baseUrl": "https://opencode.ai/zen/v1"
    }
  }
}

Compatibility

Feature	Value
supportsStore	No
supportsDeveloperRole	No
supportsReasoningEffort	Yes
supportsUsageInStreaming	Yes
maxTokensField	max_tokens
requiresToolResultName	No
requiresAssistantAfterToolResult	No
requiresThinkingAsText	No
requiresReasoningContentOnAssistantMessages	No
thinkingFormat	openai
openRouterRouting	—
vercelGatewayRouting	—
chatTemplateKwargs	—
zaiToolStream	No
supportsStrictMode	Yes
cacheControlFormat	None
sendSessionAffinityHeaders	No
supportsLongCacheRetention	Yes