API Reference

Authentication

All requests require a bearer token. Sign up at anymodl.com/account — you get $0.25 free credit instantly with no card required.

Send your key on every request:

Authorization: Bearer YOUR_ANYMODL_KEY

Quickstart

Three steps: get key → set base_url → pick any model.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_ANYMODL_KEY",
    base_url="https://api.anymodl.com/v1"
)

response = client.chat.completions.create(
    model="gpt-4o",  # or any of 605+ models
    messages=[{"role": "user", "content": "Hello, world!"}]
)
print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ANYMODL_API_KEY,
  baseURL: "https://api.anymodl.com/v1"
});

const response = await client.chat.completions.create({
  model: "claude-opus-4-5",  // or any of 605+ models
  messages: [{ role: "user", content: "Hello, world!" }]
});
console.log(response.choices[0].message.content);

curl https://api.anymodl.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_ANYMODL_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-2.0-flash",
    "messages": [{"role": "user", "content": "Hello, world!"}]
  }'

Chat Completions

OpenAI-compatible chat endpoint. Every parameter the OpenAI SDK sends is accepted.

Request parameters

Response follows the OpenAI ChatCompletion schema — choices[0].message.content contains the text.

Streaming

Set stream: true to receive tokens via server-sent events:

data: {"choices":[{"delta":{"content":"Hello"},"index":0}]}

data: {"choices":[{"delta":{"content":", world!"},"index":0}]}

data: [DONE]

Credit balance header

Every response includes your remaining balance:

Images

Generate images from text prompts. Use FLUX model IDs. FLUX.1-schnell is fast and economical.

Response: {"data": [{"url": "https://..."}]}. Download and store images — URLs are temporary.

Embeddings

Convert text to 1024-dimensional multilingual vectors for RAG, semantic search, clustering, and memory.

response = client.embeddings.create(
    model="multilingual-e5-large",
    input=["first document", "second document"]
)
vector = response.data[0].embedding  # list of 1024 floats

MCP Server

anymodl exposes a Model Context Protocol (MCP) server. Claude Desktop, Claude Code, Cursor, and any MCP-compatible runtime can attach it with one config line.

Transport: streamable-http. Authenticate with the same bearer key as the REST API.

Claude Desktop / Claude Code configuration

{
  "mcpServers": {
    "anymodl": {
      "transport": "streamable-http",
      "url": "https://mcp.anymodl.com/mcp",
      "headers": { "Authorization": "Bearer YOUR_KEY" }
    }
  }
}

Available tools (12)

Rate Limits

Rate limits apply per API key and scale when your account has $5+ in deposits.

Handling 429 errors

On rate limit, use exponential backoff starting at 1 second. Double the delay on each retry, up to ~60 seconds. Contact hello@anymodl.com for higher limits.

Error Codes

Standard HTTP status codes. Error responses include {"error": {"message": "...", "type": "...", "code": N}}.

Models

Browse the full 605+ model catalog at anymodl.com/models. Includes model IDs, makers, and pricing. New models added continuously.

curl https://api.anymodl.com/v1/models \
  -H "Authorization: Bearer YOUR_ANYMODL_KEY"