apiany-integration

name	apiany-integration
description	Use when a user asks to integrate with APIAny, choose or compare APIAny models, estimate APIAny credits or pricing, migrate OpenAI-compatible clients to APIAny, generate APIAny model usage examples, or run confirmed APIAny image and video tasks through the APIAny MCP server.

APIAny Integration Skill

Use this skill when a user asks an agent to integrate with APIAny, choose an APIAny model, compare APIAny pricing, migrate an OpenAI-compatible client to APIAny, or estimate APIAny credits.

Core Rules

• Prefer APIAny's OpenAI-compatible base URL: https://apiany.ai/v1.
• Never reveal or log API keys. Use placeholders such as <APIANY_API_KEY>.
• For model discovery, prefer the APIAny MCP server tools when available:
  • list_models
  • search_models
  • get_model
  • get_model_usage
  • estimate_cost
  • get_integration_examples
  • get_docs_context
  • create_image_task
  • create_video_task
  • get_task_status
• If MCP is not available, use public docs and https://apiany.ai/api/v1/models.
• Treat pricing as an estimate unless returned by the live API at request time.
• For image/video/audio generation, explain async task behavior: create task, poll status or use callback, charge on success, refund on failure.
• Do not call paid generation endpoints unless the user explicitly asks, provides an API key, and confirms the paid request. For MCP tools, pass confirm_paid_request=true only after confirmation.

Workflow

1. Understand the user's target workflow:
   • Chat / text
   • Image generation or editing
   • Video generation
   • Audio or music
   • Cost comparison
   • Migration from OpenAI-compatible client
2. Select candidate models:
   • Search by modality, provider, model family, or capability.
   • Compare model pricing and capabilities.
   • Prefer stable public model ids.
3. Provide integration code:
   • Use get_model_usage for the selected model whenever the MCP server is available.
   • Use the OpenAI-compatible SDK for chat models when applicable.
   • Use APIAny media endpoints for image/video models and explain task polling.
   • Set base_url / baseURL to https://apiany.ai/v1.
   • Set API key from environment, not inline.
4. Estimate cost:
   • Use estimate_cost when available.
   • Mention that server-side billing rules are authoritative.
5. Provide verification steps:
   • Check API key.
   • Run a small request.
   • Inspect task status or response.
   • Confirm credit usage in dashboard.

Paid Media Workflow

1. Use get_model_usage to show the selected model's endpoint, payload, and expected async behavior.
2. Ask for confirmation before a paid task.
3. Call create_image_task or create_video_task with confirm_paid_request=true.
4. Poll get_task_status until the task succeeds or fails.
5. Report task_id, status, result links, charged credits, and any error.

OpenAI-Compatible Examples

Python:

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://apiany.ai/v1",
    api_key=os.environ["APIANY_API_KEY"],
)

response = client.chat.completions.create(
    model="gpt-5.5",
    messages=[{"role": "user", "content": "Write a launch checklist."}],
)

print(response.choices[0].message.content)

JavaScript:

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://apiany.ai/v1",
  apiKey: process.env.APIANY_API_KEY,
});

const response = await client.chat.completions.create({
  model: "gpt-5.5",
  messages: [{ role: "user", content: "Write a launch checklist." }],
});

console.log(response.choices[0].message.content);

Suggested Agent Responses

When asked to choose a model:

1. Search APIAny models.
2. Return 2-4 candidates.
3. Explain tradeoffs: modality, quality, price, speed, provider.
4. Include model ids and a usage example from get_model_usage.

When asked to migrate from OpenAI:

1. Keep the existing SDK if possible.
2. Change only baseURL and API key.
3. Replace model id with an APIAny public model id.
4. Run a minimal smoke test.

Required User Inputs

Ask the user for:

• Target modality and desired output.
• Budget or latency preference.
• API key only when they want to run paid requests.
• Preferred language/framework for examples.