# Amux AI vs ZenMux: more native protocols, fuller team controls

> Amux AI and ZenMux both offer a unified LLM API priced per model and provider. The differences: Amux AI offers low-cost channels far below official pricing plus discounted official channels and natively supports the generateContent protocol that the Gemini SDK uses by default, plus Alibaba Cloud Model Studio, Volcengine Seedance, MiniMax and xAI image and video protocols; it offers enterprise-grade organization management with workspaces, a finance role and three levels of spend limits; bodies of successful text calls are not stored; and the cost of every call is available within seconds. ZenMux offers AI insurance payouts, coding-assistant subscription plans, and embeddings and audio endpoints.

Last verified 2026-10-19 · https://amux.ai/vs/zenmux

- **90%** Max saving vs. official price: From low-cost channels, based on live catalog prices; stable official channels are discounted too
- **4** Native video vendor protocols: Model Studio, Seedance, MiniMax and xAI — existing code only changes the base URL
- **3 levels** Team spend limits: Per workspace, member and key; ZenMux organizations have only admin and member roles
- **Seconds** To see per-call cost: ZenMux billing data is queried via a management API 3–5 minutes after the request

## Amux AI vs ZenMux, feature by feature

| Feature | Amux AI | ZenMux |
|---|---|---|
| Inference pricing | No inference markup; low-cost channels go far below official pricing, and stable official channels are discounted too | Priced per model and provider, with limited-time discounts on some models |
| Text protocols | Native OpenAI Chat, OpenAI Responses, Anthropic Messages and Gemini (generateContent, the Gemini SDK default), with most models callable across protocols | Four protocols, with Gemini in Vertex AI path form (requires vertexai=True); Anthropic protocol support is still being rolled out model by model |
| Image and video | Image: OpenAI Images and Alibaba Cloud Model Studio. Video: native Model Studio, Volcengine Seedance, MiniMax and xAI request formats plus a unified task API — existing client code only needs a new base URL and key | Images through OpenAI Images or the Vertex AI protocol; video through the Vertex AI async protocol or a proprietary video API |
| Embeddings and audio | Embeddings, speech and rerank endpoints (Coming soon) | Embeddings, text-to-speech, speech-to-text and rerank |
| Model coverage | 70+ text, image and video models, with more added regularly; 52 models can switch between providers | Its model catalog lists 240+ models |
| Routing and failover | Five routing strategies, configurable per API key; automatic failover on connection errors, timeouts, rate limits and server errors, with failed attempts not billed | Sorted by time to first token by default; provider pinning, sorting by price or throughput, and model fallbacks |
| Teams and budgets | Enterprise-grade organization management: organizations, workspaces and roles (including a finance role); spend limits per workspace, member and key | Organizations with admin and member roles and per-member limits; up to 50 members per organization |
| Data privacy | Request and response bodies of successful text calls are not stored; customer content is never used for training | Request and response content stored while data services are on, zero retention when off; the privacy policy reserves the right to use anonymized content to improve models |
| Usage and cost | Itemized charges for every call, visible in the logs seconds after the call ends; monthly statements exported to Excel by key, model and workspace | Billing data queried via a management API 3–5 minutes after the request; usage and cost pages exclude subscription usage |
| Subscription plans | Pay as you go; subscription plans (Coming soon) | Coding-assistant subscriptions at $20–$200/month, limited by the terms to personal use and not for production |
| Failure compensation | Failed attempts are not billed; the platform absorbs the upstream cost | AI insurance checked daily, with eligible payouts credited the next day |

## Choose Amux AI if

- You want leading models such as Claude, GPT and Gemini far below official pricing through low-cost channels, or at a discount through stable official channels, with failed attempts left off your bill
- You use the Google Gemini SDK in its default form, or want to call image and video models in the native Alibaba Cloud, Volcengine, MiniMax or xAI request formats
- Your team needs budgets per workspace, member and key, with a finance role managing top-ups and invoices
- You want bodies of successful text calls left unstored and the cost of every call visible within seconds

## Choose ZenMux if

- You value a daily-checked compensation scheme such as AI insurance
- You use a coding assistant personally and want a fixed monthly subscription
- You need embeddings, speech or rerank endpoints today (on the Amux AI roadmap)
- You need more long-tail models

## Switch from ZenMux to Amux AI

Both work with the OpenAI and Anthropic SDKs, use the same manufacturer/model ID format and the same provider suffix syntax. Switching only changes the endpoint and key — your application code stays as it is.

1. Sign up for Amux AI and create an API key in the console.
2. Replace the base URL with the Amux AI endpoint and the API key with your new key.
3. Keep your existing model IDs; for the few models whose version format differs, copy the ID from the model page.

- OpenAI SDK base URL: `https://gateway.amux.ai/v1`
- Anthropic SDK / Claude Code base URL: `https://gateway.amux.ai`

```python
import os
from openai import OpenAI

client = OpenAI(base_url="https://gateway.amux.ai/v1", api_key=os.environ["AMUX_API_KEY"])
response = client.chat.completions.create(
    model="openai/gpt-6.1-sol",
    messages=[{"role": "user", "content": "Hello"}],
)
```

- To pin a provider, append ":provider" to the model ID as before — check the provider slug on the Amux AI model page — or set a routing strategy on the API key.
- The Gemini SDK does not need vertexai=True; call generateContent in its default form. Claude Code and the Anthropic SDK use the endpoint without /v1.

## Frequently asked questions

### What is the main difference between Amux AI and ZenMux?

Both price per model and provider. Amux AI offers low-cost channels far below official pricing plus discounted official channels, natively supports the default form of the Gemini SDK plus vendor image and video protocols, offers enterprise-grade organization management with workspaces, a finance role and three levels of spend limits, does not store bodies of successful text calls, and shows per-call cost within seconds. ZenMux offers AI insurance, coding-assistant subscriptions, and embeddings and audio endpoints.

### How can Amux AI price models below the official rate?

Amux AI connects each model to several providers in two kinds of channels: low-cost channels priced far below official pricing for cost-sensitive workloads, and official channels from model makers and major clouds that are stable and also discounted. Live per-provider prices are published on each model page; you can restrict an API key or workspace to official channels only, or set price-first routing.

### How does calling through the Google Gemini SDK differ?

Amux AI implements the generateContent protocol that the Gemini SDK uses by default, so you only change the endpoint and key. ZenMux exposes Gemini in Vertex AI path form, which requires vertexai=True in the SDK.

### What code changes are needed to switch from ZenMux to Amux AI?

Usually two: point the base URL at the Amux AI endpoint and replace the API key with an Amux AI key. Model IDs and the ":provider" suffix use the same syntax; check the Amux AI model page for exact model names and provider slugs.

### Does Amux AI store my prompts and responses?

Request and response bodies of successful text calls are not stored, and customer content is never used for training; prompts for image and video tasks are kept with the task for 30 days. Failed calls keep a limited sample for troubleshooting for up to 15 days, which you can turn off with a request header.

### Can Claude Code, Cursor and similar tools switch from ZenMux to Amux AI?

Yes. Amux AI natively implements the Anthropic Messages and OpenAI protocols, and setup guides cover Claude Code, Codex, Gemini CLI, Cursor, Cline, Cherry Studio and more — replace the endpoint and key as described.

### Can I use ZenMux and Amux AI together?

Yes. Both work with the OpenAI SDK, so you can split traffic by model or use case — for example, send mainstream models and video generation through Amux AI and keep a personal coding assistant on a ZenMux subscription.

_ZenMux and other product names mentioned on this page are trademarks of their respective owners. Information about third-party products is compiled from their official public sources for evaluation purposes only and may change; refer to each vendor's website for current details. To report an inaccuracy, contact support@amux.ai._