⌁ Claude · GPT · Gemini · DeepSeek — one endpoint

One API key for every AI model you use.

OpenMux is an OpenAI-compatible AI gateway. Route requests to any model, track usage per key, and manage billing — all from a single control plane.

OpenAI-compatible Transparent usage logs PAYG & subscription keys No vendor lock-in
⚡ Gateway Console api.openmux.ai
Smart Routing 200 OK
POST /v1/chat/completions
claude-sonnet-4-6
Anthropic
Primary
gpt-4o
OpenAI
Fallback
deepseek-reasoner
DeepSeek
Standby
p95 latency218ms
uptime99.98%
spend this month
$17.40
real-time
Active models7
Providers4
Key modesPAYG + Sub

A unified AI API layer for developers and teams.

Keep your existing SDK. Swap the base URL and API key — OpenMux handles routing, billing, and observability across every provider.

🔑
Subscription Keys

Stable monthly quota for Claude Code, Cursor, and Codex workflows.

💳
Pay-as-you-go

Top up a balance and pay only for what you use, per token.

📊
Usage Logs

Per-request breakdown by model, token count, cost, and status.

🧩
OpenAI-compatible

Drop-in replacement — no SDK changes, no rewrites required.

Send a request. OpenMux routes it to the right model.

From solo developer tools to production services — all your keys, models, spend, and logs in one console.

1
Sign in with Google, Apple, or email
2
Add credit or choose a subscription plan
3
Create a key and swap your base URL

Two-minute integration. No rewrites.

OpenAI-compatible protocol means you only change the base URL and API key — everything else stays the same.

Claude Code

Use a subscription key for high-frequency terminal coding.

Cursor

Paste the OpenAI-compatible base URL in settings.

Codex

Set OPENAI_BASE_URL to route to Claude or DeepSeek.

Cline

Plug into agentic editor workflows with any model.

quickstart.py
# pip install openai
from openai import OpenAI

client = OpenAI(
    base_url="https://api.openmux.ai/v1",
    api_key="om-sk-your-key",
)

response = client.chat.completions.create(
    model="claude-sonnet-4-6",
    messages=[{
        "role": "user",
        "content": "Ship it.",
    }],
)

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

Every model. Visible pricing. Faster decisions.

Pick the right model for each job — high-capability, cost-efficient, or reasoning — and assign different keys to different teams or projects.

Claude Sonnet 4.6 Recommended
claude-sonnet-4-6
Anthropic$3 / $15 per M tok
Claude Opus 4.5 Most Capable
claude-opus-4-5
Anthropic$15 / $75 per M tok
GPT-4o Versatile
gpt-4o
OpenAI$2.5 / $10 per M tok
DeepSeek R1 Reasoning
deepseek-reasoner
DeepSeek$0.55 / $2.19 per M tok