MCP 服务器

APIAny MCP Server

Provides tools for discovering AI models, checking pricing, retrieving API documentation context, and generating integration examples; also supports paid async image/video tasks.

README

APIAny MCP

Official APIAny MCP server and agent skill package for model discovery, pricing lookup, documentation context, and OpenAI-compatible integration examples.

Official website: https://apiany.ai

What Is Included

src/server.js: MCP server for APIAny public model discovery, docs context, usage examples, and confirmed paid media tasks.
skills/apiany-integration: Agent skill for choosing APIAny models, estimating credits, and generating safe integration code.
examples/mcp.json: MCP client configuration example.
docs/distribution.md: Release checklist for npm, MCP registries, and skill marketplaces.

Install MCP

npx @apiany/mcp

During local development from this repository:

npm install
npm start

Configuration

Environment variables:

APIANY_BASE_URL: APIAny site/API base URL. Defaults to https://apiany.ai.
APIANY_API_KEY: optional APIAny API key. Read-only public tools work without it.

Claude Code / Cursor / VS Code Config Example

{
  "mcpServers": {
    "apiany": {
      "command": "npx",
      "args": ["-y", "@apiany/mcp"],
      "env": {
        "APIANY_BASE_URL": "https://apiany.ai"
      }
    }
  }
}

Tools

list_models: list public APIAny models with pricing and capability metadata.
search_models: search models by text, provider, modality, or capability.
get_model: get one public model by model id or display name.
estimate_cost: estimate credits from public pricing metadata.
get_integration_examples: return cURL, Python, JavaScript, Go, Java, or PHP examples for chat, image, or video models.
get_model_usage: return endpoint, payload, async behavior, and language example for one or more models.
get_docs_context: return compact APIAny documentation context from /llms.txt.
create_image_task: create a paid async image task. Requires APIANY_API_KEY and confirm_paid_request=true.
create_video_task: create a paid async video task. Requires APIANY_API_KEY and confirm_paid_request=true.
get_task_status: read async task status. Requires APIANY_API_KEY.

Agent Skill

The APIAny Integration skill lives in skills/apiany-integration.

It helps agents:

Choose APIAny models by modality, provider, pricing, and capability.
Generate examples for chat, image, video, and media workflows.
Explain async task creation and polling.
Require explicit confirmation before paid generation tasks.

Prompt Examples

Ask an agent:

Use the APIAny MCP server to find an image model for product photos, compare pricing, and show a JavaScript integration example.

Use APIAny MCP to estimate the credits for 20 requests to gpt-5.5 with 10k input tokens and 5k output tokens.

Use APIAny MCP to show JavaScript usage for nano-banana-pro and Python usage for veo3-1-fast.

Use APIAny MCP to create an image task with nano-banana-pro after I confirm the paid request, then poll the task status.

Safety

Read-only tools work without an API key. Paid generation tools require APIANY_API_KEY and confirm_paid_request=true; the server does not persist API keys.