All-in-One LLM API Provider

One API Key, Every LLM Model

The most affordable LLM API provider for OpenClaw, Hermes Agent, and AI agent developers. Access Claude API, GPT-4o, Gemini and more through native API formats — with discounted pricing and balance that never expires.

Documentation

New users will receive a free trial balance after creating an API key — no credit card required

Why Developers Choose Our LLM API

Built for AI agents like OpenClaw and Hermes Agent. One LLM API key unlocks all leading models from every major provider.

🔗

All Models, One API Key

Claude API key, GPT-4o, Gemini — access every major LLM provider through a single API key. No need for multiple accounts or separate billing.

💰

Discounted LLM Pricing

Up to 90% off official pricing on every model. The cheapest way to access Claude, GPT-4o and Gemini APIs as your LLM provider.

♾️

Balance Never Expires

No monthly subscriptions or expiring credits. Top up your LLM API balance once — it never expires. Pay only for the tokens you actually use.

🔌

Native API Formats

OpenAI Completions, Anthropic Messages, Gemini — each LLM API uses its native format. Drop-in replacement for OpenClaw, Hermes Agent and any AI agent.

The Best LLM API Provider for AI Agents

Ideal LLM Provider for Hermes Agent

Hermes Agent is a personal AI assistant that supports multiple LLM providers through its flexible configuration system. Our LLM API integrates seamlessly — configure base_url and api_key in ~/.hermes/config.yaml to point to our endpoint, and Hermes Agent instantly gains access to all models. Whether you prefer Claude for reasoning tasks, GPT-4o for general conversations, or Gemini for cost-efficient operations, our agent API provides them all. The Hermes Agent API setup takes less than a minute with our LLM provider.

Perfect for OpenClaw API Integration

OpenClaw is a powerful open-source AI coding agent that relies on external LLM APIs for intelligence. Our LLM API provider works as a drop-in replacement in your OpenClaw configuration — just set the base URL and your API key. You get access to Claude, GPT-4o, Gemini and dozens more models through a single OpenClaw API provider config. No need to juggle multiple Claude API keys or OpenAI accounts. Our LLM provider handles the routing, load balancing, and billing so you can focus on building with OpenClaw.

Save Money with Aggregated LLM API Pricing

Managing separate accounts with Anthropic, OpenAI, and Google means separate billing, separate API keys, and separate balance tracking. Our all-in-one LLM API provider consolidates everything into a single balance. You get discounted pricing — up to 90% off official rates — because we aggregate demand across thousands of developers. For startups and individual developers building with OpenClaw or Hermes Agent, this means significant savings on your monthly LLM API costs.

Enterprise-Grade Reliability for Your Agent API

Your AI agent needs reliable LLM API access. Our provider features automatic model fallback — if one provider has issues, requests seamlessly route to backup providers. Combined with multiple API key rotation for load balancing, your OpenClaw and Hermes Agent deployments stay responsive even during provider outages. All LLM API calls are metered transparently with per-request usage tracking visible in your dashboard.

Available LLM Models & Pricing

Transparent per-token LLM API pricing. All models accessible through a single API key.

Model	Type	Provider	API Format	Input	Output	Discount

No models found.

LLM API prices per 1M tokens in USD. Discount applied on top of official model pricing from each provider.

LLM API Use Cases

Our LLM provider powers thousands of AI agents, chatbots, and developer tools worldwide.

🤖

AI Coding Agents

Power your OpenClaw setup with the best LLM API provider. Route coding tasks to Claude Sonnet for complex reasoning and GPT-4o for quick edits — all through one API key with unified billing.

💬

Personal AI Assistants

Configure Hermes Agent or any personal assistant with our LLM API. Switch between models on the fly — use affordable Gemini Flash for simple queries and premium Claude for deep analysis, all from one provider.

🏢

Production Applications

Bonus +4%

Credited to balance: $41.60

Balance never expires

Access all LLM models

Newcomer

Automatically top up your balance when it falls below a threshold. Your card will be charged in the background — no interruption to your API calls.

Securely save your card via Stripe to enable automatic charging.

Card on file

Enable Auto Top-up

Top up when balance falls below

$

Top-up amount

You will receive the credited balance shown on each plan.

Auto recharge is not enabled

🔒 Your card details are securely handled by Stripe and never stored on our servers.

Get Started with Our LLM API in 3 Steps

1

Create Your API Key

Sign up and create your LLM API key in the dashboard. Each key works with all supported models — Claude, GPT-4o, Gemini and more.

2

Top Up Your Balance

Choose a plan and top up. Your LLM API balance never expires. Newcomers can try the Lite plan at $9.99 with no bonus — perfect for testing.

3

Connect Your Agent

Set the base URL and API key in OpenClaw, Hermes Agent, or any OpenAI-compatible client. Start making LLM API calls immediately.

Quick Start — Use LLM API in 30 Seconds

cURL

curl https://banana2.pro/llm-api/v1/chat/completions \
  -H "Authorization: Bearer llm_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": true
  }'curl https://banana2.pro/llm-api/anthropic/v1/messages \
  -H "x-api-key: llm_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": true
  }'curl "https://banana2.pro/llm-api/v1beta/models/gemini-2.5-flash:streamGenerateContent" \
  -H "Authorization: Bearer llm_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{"parts": [{"text": "Hello!"}]}]
  }'curl https://banana2.pro/llm-api/v1/responses \
  -H "Authorization: Bearer llm_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "input": "Hello!",
    "stream": true
  }'

Works instantly with OpenClaw, Hermes Agent, and any OpenAI/Anthropic/Gemini-compatible AI agent client.

POST /llm-api/v1/images/generations cURL

# 1. Submit image generation task
curl -X POST https://banana2.pro/llm-api/v1/images/generations \
  -H "Authorization: Bearer llm_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "A futuristic cityscape at sunset",
    "size": "16:9",
    "resolution": "1k"
  }'

# 2. Poll until done
curl https://banana2.pro/llm-api/v1/tasks/{task_id} \
  -H "Authorization: Bearer llm_YOUR_API_KEY"

gpt-image-2 Python

import requests, time

BASE = "https://banana2.pro/llm-api"
HEADERS = {
    "Authorization": "Bearer llm_YOUR_API_KEY",
    "Content-Type": "application/json"
}

resp = requests.post(f"{BASE}/v1/images/generations", headers=HEADERS, json={
    "model": "gpt-image-2",
    "prompt": "A futuristic cityscape at sunset",
    "size": "16:9",
    "resolution": "1k"
})
task_id = resp.json()["task_id"]

while True:
    time.sleep(5)
    r = requests.get(f"{BASE}/v1/tasks/{task_id}", headers=HEADERS).json()
    if r["status"] == "completed":
        print(r["result"]["urls"])
        break
    elif r["status"] == "failed":
        print("Error:", r.get("error"))
        break

Image generation is async — submit a task and poll for results. The charge is deducted upfront; refunded automatically if the task fails.

POST /llm-api/v1/images/generations cURL

# 1. Submit task with webhook_url
curl -X POST https://banana2.pro/llm-api/v1/images/generations \
  -H "Authorization: Bearer llm_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "A futuristic cityscape at sunset",
    "size": "16:9",
    "resolution": "1k",
    "webhook_url": "https://your-server.com/webhook/images"
  }'

# 2. Platform pushes result to your server when done
{
  "task_id": "media_a1b2c3d4e5f6g7h8",
  "status": "completed",
  "model": "gpt-image-2",
  "created_at": 1735000000.0,
  "result": {
    "urls": ["https://files.example.com/output.png"]
  }
}

# Your server receives the webhook callback (Flask example) Python

from flask import Flask, request, jsonify

app = Flask(__name__)

@app.route("/webhook/images", methods=["POST"])
def handle_webhook():
    data = request.json
    task_id = data["task_id"]
    status  = data["status"]

    if status == "completed":
        urls = data["result"]["urls"]
        print(f"Task {task_id} done: {urls}")
    elif status == "failed":
        print(f"Task {task_id} failed: {data.get('error')}")

    return jsonify({"ok": True})

Webhook mode requires no polling — the platform automatically POSTs the result to your server when done. Recommended for production.

Our LLM Provider vs Direct API Access

See why developers prefer our aggregated LLM API provider over managing multiple direct accounts.

Feature	Banana 2 Pro	Direct Access
Number of accounts needed	1 account	3+ accounts
Billing management	Single balance	Multiple invoices
Balance expiration	Never expires	Varies by provider
Model access	All models, one key	One provider per key
Pricing vs official	Up to 90% off	Full price
Automatic failover	Built-in	Build your own
OpenClaw / Hermes Agent	Drop-in config	Per-provider setup

Frequently Asked Questions about Our LLM API

No. Your LLM API balance never expires. Top up once and use it whenever you need — no monthly subscriptions, no expiring credits. This makes our LLM provider ideal for both active and occasional users.

We support all four major LLM API formats: OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, and Gemini. Each model uses its native API format directly — no translation layer needed.

Simply add our LLM API as a custom provider in your OpenClaw configuration with your API key and our base URL. OpenClaw will route requests through our provider to access Claude, GPT-4o, and all other models. See our documentation for the exact OpenClaw API config.

Set OPENAI_API_KEY to your LLM API key and OPENAI_BASE_URL to our endpoint in Hermes Agent's .env file. Hermes Agent will use our LLM provider for all model requests. The agent API setup takes under a minute.

Yes. Our LLM API provider offers up to 90% off official pricing. Plus, you get access to all models with a single balance — no need to manage multiple Claude API keys, OpenAI accounts, or Google API billing.

Each account can create up to 5 active LLM API keys. Contact us if you need more for your AI agent setup.

Our LLM API provider supports automatic model fallback. If the primary provider for a model is experiencing issues, requests can be routed to a backup provider to ensure your agent API calls succeed.

Yes! We offer redeem codes for promotional balance. Enter your code in the Dashboard to instantly add the balance to your LLM API account.