# LLM4Agents

> Unified LLM inference platform for AI agents. OpenAI-compatible API. Pay-per-token with USDT/USDC on Solana and Polygon.

## What is this

LLM4Agents is an LLM proxy API designed for autonomous AI agents. Agents register themselves, generate a crypto wallet, deposit stablecoins, and make inference calls to 345+ LLMs using the standard OpenAI API format — no human required at any step.

## Who is this for

- AI agents that need to call LLMs programmatically without a human credit card
- Multi-agent systems that require isolated billing per agent
- Developers building autonomous agents that self-fund their own inference

## How to integrate (as an agent)

### Step 1 — Register (no auth required)

POST https://api.llm4agents.com/api/v1/agents/register
Content-Type: application/json

{"name": "your_agent_name"}

Response: { "uuid": "...", "apiKey": "sk-proxy-..." }

Rate limit: 5 registrations per IP per hour.

### Step 2 — Generate a deposit wallet

POST https://api.llm4agents.com/api/v1/wallets/generate
Authorization: Bearer sk-proxy-...
Content-Type: application/json

{"chain": "solana", "token": "USDT"}

Supported chains: solana, polygon
Supported tokens: USDT, USDC

### Step 3 — Deposit USDT or USDC

Send USDT or USDC to the wallet address returned in Step 2. Balance is credited automatically on-chain confirmation. No human approval needed.

### Step 4 — Call any LLM

POST https://api.llm4agents.com/v1/chat/completions
Authorization: Bearer sk-proxy-...
Content-Type: application/json

{
  "model": "anthropic/claude-3-haiku",
  "messages": [{"role": "user", "content": "Hello"}]
}

This is 100% OpenAI SDK compatible. Change only base_url to https://api.llm4agents.com.

Each response includes cost headers:
- X-Cost-Usd-Cents: cost of this call
- X-Balance-Remaining-Cents: remaining balance

### Check balance

GET https://api.llm4agents.com/api/v1/balance
Authorization: Bearer sk-proxy-...

Response: { "availableUsd": "...", "totalDepositedUsd": "...", "totalSpentUsd": "..." }

## Rate limits

- Registration: 5 req / hour / IP
- Chat completions: 600 req / min / API key
- Other endpoints: 120 req / min / API key

Returns 429 on limit. Use exponential backoff.

## Available models

345+ models from OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, xAI, Qwen, Cohere, Perplexity, Amazon, NVIDIA and more.

Full list: GET https://api.llm4agents.com/api/v1/models (no auth required)
Full OpenAPI spec: https://api.llm4agents.com/docs/openapi.json

## Key facts

- No subscription, no monthly fee — pay only for tokens consumed
- Each agent has its own isolated wallet and API key
- Every transaction has a txHash verifiable on-chain
- Webhooks notify on deposit: POST /api/v1/webhooks/deposit
- Transaction history: GET /api/v1/transactions