airpx — Multi-provider AI Gateway

Overview

airpx is a multi-provider AI gateway that exposes 330+ models from 30 providers through a single, unified API. It is fully compatible with both the OpenAI and Anthropic APIs, so you can use it as a drop-in replacement in any tool or SDK that supports those formats.

Key features: smart routing across multiple provider keys, automatic failover, per-model pricing with token-level billing, streaming with SSE, tool/function calling, thinking/reasoning support, and a growing catalog of free models.

Base URL (OpenAI-compatible): https://airpx.cc/v1 — Base URL (Anthropic-compatible): https://airpx.cc

Quick Start

Get up and running in three steps.

1. Sign up at https://airpx.cc and create an API key in the dashboard. 2. Set your base URL to https://airpx.cc/v1 (OpenAI) or https://airpx.cc (Anthropic). 3. Send your first request:

curl

curl https://airpx.cc/v1/chat/completions \
  -H "x-api-key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "grok-4-reason",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

That's it! Many models including grok-4-reason are free — you can start immediately after signing up.

Authentication

All /v1/* proxy endpoints accept your API key via one of two headers.

Pass your proxy key (starts with sk-proxy-*) in either header: • x-api-key: YOUR_KEY • Authorization: Bearer YOUR_KEY Both headers are interchangeable. Use whichever your tool or SDK supports.

Important: proxy keys (sk-proxy-*) are used for /v1/* endpoints only. They are NOT the same as dashboard login tokens.

curl https://airpx.cc/v1/chat/completions \
  -H "x-api-key: sk-proxy-abc123" \
  -H "Content-Type: application/json" \
  -d '{"model": "grok-4-reason", "messages": [{"role": "user", "content": "Hi"}]}'

API Reference

Public proxy endpoints. All require a valid API key.

POST/v1/chat/completionsChat completion (OpenAI format). Streaming supported.

POST/v1/messagesAnthropic Messages API (native format). Streaming supported.

POST/v1/responsesOpenAI Responses API (Codex CLI). Streaming supported.

GET/v1/modelsModel catalog with pricing and capabilities.

POST/v1/embeddingsText embedding generation (OpenAI format).

POST/v1/audio/transcriptionsAudio transcription (Whisper-compatible).

All endpoints accept JSON request bodies and return JSON responses. For streaming, set "stream": true and receive Server-Sent Events (SSE).

Integrations

airpx works as a drop-in backend for popular coding assistants and AI tools.

🔧Claude Code

Add environment variables to ~/.bashrc or ~/.zshrc
Run claude as usual

bash

export ANTHROPIC_BASE_URL=https://airpx.cc
export ANTHROPIC_API_KEY=YOUR_KEY

# Then just run:
claude

🔧Pi Coding Agent

Set Anthropic environment variables
Run pi as usual — uses /v1/messages with streaming SSE

bash

export ANTHROPIC_BASE_URL=https://airpx.cc
export ANTHROPIC_API_KEY=YOUR_KEY

# Or in .env / config:
ANTHROPIC_BASE_URL=https://airpx.cc
ANTHROPIC_API_KEY=YOUR_KEY

🔧Cursor

Open Settings → Models → OpenAI API Key + Base URL
Enter your airpx key and base URL

text

API Key: YOUR_KEY
Base URL: https://airpx.cc/v1

🔧Cline (VS Code)

Open extension settings → API Provider → OpenAI Compatible
Enter your airpx key and base URL

text

Base URL: https://airpx.cc/v1
API Key: YOUR_KEY
Model: claude-opus-4

🔧Aider

Set OPENAI_BASE_URL environment variable
Run aider with any supported model

bash

export OPENAI_BASE_URL=https://airpx.cc/v1
export OPENAI_API_KEY=YOUR_KEY

aider --model claude-opus-4

🔧OpenCode

Set OPENAI_BASE_URL environment variable
Run opencode

bash

export OPENAI_BASE_URL=https://airpx.cc/v1
export OPENAI_API_KEY=YOUR_KEY

opencode

🔧Codex CLI

Set openai_base_url in ~/.codex/config.toml
Export OPENAI_API_KEY and run codex

openai_base_url = "https://airpx.cc/v1"

SDKs

Use standard OpenAI and Anthropic SDKs — just point them at airpx.

pip install openai

from openai import OpenAI

client = OpenAI(
    base_url="https://airpx.cc/v1",
    api_key="YOUR_KEY"
)

r = client.chat.completions.create(
    model="grok-4-reason",  # free
    messages=[{"role": "user", "content": "Hello!"}]
)
print(r.choices[0].message.content)

Models

airpx exposes 330+ models across 30 providers. Browse the full catalog at /v1/models.

The model catalog is available at GET /v1/models and returns all active models with their pricing, capabilities, and provider information.

Models are split into free and paid tiers. Many models are available for free — check GET /v1/models for the current list (free models show $0 pricing). Paid models are billed per 1M tokens with separate rates for input, output, cache read, and cache write tokens. Example pricing: grok-4-fast $0.20/$0.50, gpt-5.5 $5/$30, claude-opus-4 $5/$25 per 1M tokens (input/output).

List models

curl https://airpx.cc/v1/models \
  -H "x-api-key: YOUR_KEY"

Billing

Prepaid balance with token-level billing.

airpx uses a prepaid balance model. You top up your balance in the dashboard, and each request deducts the cost based on actual token usage and the model's pricing.

Billing is calculated per request based on: • Input tokens × model input price (per 1M tokens) • Output tokens × model output price (per 1M tokens) • Cache read tokens × cache read price • Cache write tokens × cache write price Your balance is checked before each request. If insufficient, the request returns HTTP 402 (Payment Required).

Free models do not deduct from your balance and can be used without topping up. Check the Models page or GET /v1/models for the current free model list.

Streaming

Both OpenAI and Anthropic formats support real-time streaming via Server-Sent Events (SSE).

Set "stream": true in your request body to receive incremental responses. The proxy streams directly from the upstream provider, preserving the original SSE format for both OpenAI and Anthropic protocols.

{
  "model": "grok-4-reason",
  "stream": true,
  "messages": [{"role": "user", "content": "Hello!"}]
}

Thinking / reasoning is also supported. Use reasoning_effort (OpenAI style) or thinking: {type: "enabled", budget_tokens: N} (Anthropic style). Responses include reasoning_content or thinking blocks accordingly.

{
  "model": "claude-opus-4",
  "reasoning_effort": "high",
  "stream": true,
  "messages": [{"role": "user", "content": "Solve this step by step..."}]
}

Bring Your Own Key

Use your own provider API keys for direct access and custom billing.

If you have your own API key from a provider (e.g., OpenAI, Anthropic, Google), you can configure it in the dashboard. Requests routed through your own key are billed directly by the provider — airpx acts as a transparent proxy.

BYOK is useful when you need higher rate limits, specific model access, or prefer to use your existing provider agreement. Your own keys are never shared with other users.

Limits & Errors

HTTP status codes and common error scenarios.

The proxy returns standard HTTP status codes: • 401 Unauthorized — invalid or missing API key • 402 Payment Required — insufficient balance for the requested model • 403 Forbidden — model not allowed for your key or account tier • 429 Too Many Requests — rate limit exceeded, retry after backoff • 500 Internal Server Error — unexpected proxy error • 502 Bad Gateway — upstream provider error or timeout

Rate limits depend on your account tier and key configuration. If you hit a 429, wait a few seconds and retry. Balance errors (402) require topping up in the dashboard.

FAQ

Is airpx OpenAI-compatible?▼

Yes. airpx is a drop-in replacement for the OpenAI API. Point any OpenAI SDK or tool at https://airpx.cc/v1 and it will work out of the box.

Is the Anthropic Messages API supported?▼

Yes. airpx natively supports the Anthropic Messages format at POST /v1/messages. Set your Anthropic base URL to https://airpx.cc and use your airpx key.

Which auth header should I use?▼

You can use either x-api-key: YOUR_KEY or Authorization: Bearer YOUR_KEY. Both work identically for all /v1/* endpoints.

Where do I get an API key?▼

Which models are free?▼

Many models are available for free. The list changes as providers update their offerings — check GET /v1/models for the current catalog. Free models show $0 pricing.

Does streaming work?▼

Yes. Set "stream": true in your request body. Both OpenAI and Anthropic formats stream via Server-Sent Events (SSE).

Does tools / function calling work?▼

Yes. Tools and function calling are fully supported for both OpenAI and Anthropic formats. Pass your tool definitions as usual.

Can I use my own provider key (BYOK)?▼

Yes. You can add your own OpenAI, Anthropic, or other provider keys in the dashboard. Requests routed through your key are billed by the provider directly.

Is there a Swagger / OpenAPI spec?▼

Yes. Visit /v1/docs for the interactive Swagger UI with the full OpenAPI specification.

How do I contact support?▼

Open an issue on our repository or reach out via the dashboard contact form.