Discord Download

ZenMux alternatives for Codex and Claude Code

AutoRouter9 min read

Use AutoRouter, the Mac app for Codex and Claude Code, if you code with agents and already pay for a Claude or ChatGPT plan. It lets both apps share those plans, puts 20+ other models in the model picker you already use with no API key, and picks a model for each task on its own. It also works in Cursor, Claude Desktop, T3 and Terminal. ZenMux is a hosted LLM gateway: one ZenMux key reaches 100+ models, paid from ZenMux credits or a ZenMux Builder Plan. OpenRouter, Requesty, Portkey and LiteLLM are the other gateways people weigh against it.

AutoRouter wrote this post, so weigh its rows accordingly. Each claim about another product comes from that product's own docs, site or pricing page, linked where it is used and read on November 17, 2026 unless the sentence says otherwise.

What is ZenMux?

ZenMux calls itself an enterprise LLM platform with a unified API for 100+ AI models from OpenAI, Anthropic, Google, DeepSeek and others (ZenMux). It speaks the OpenAI protocols at zenmux.ai/api/v1 and the Anthropic protocol at zenmux.ai/api/anthropic, names models as provider/model, and serves requests from Cloudflare's edge network (Quickstart, introduction).

ZenMux vs AutoRouter puts the two side by side in one table.

Why look for a ZenMux alternative?

Most developers who code in Codex or Claude Code look because ZenMux adds a second bill and a key to manage. Its Claude Code guide sets ANTHROPIC_BASE_URL to ZenMux, puts a ZenMux key in ANTHROPIC_AUTH_TOKEN, blanks ANTHROPIC_API_KEY and maps other models through the ANTHROPIC_DEFAULT_*_MODEL variables in your shell profile. Its Codex guide exports ZENMUX_API_KEY and adds a ZenMux provider to config.toml. Either way, requests are paid from ZenMux credits or a Builder Plan, not from a Claude or ChatGPT plan:

None of that is a flaw for an app you ship, where one key for 100+ models is the point. It is the reason a developer who already pays for a Claude or ChatGPT plan looks elsewhere.

How is AutoRouter different from ZenMux?

AutoRouter uses the plans you already pay for first and picks a model for each task inside your agent, with no keys or config files to edit. AutoRouter, the Mac app for Codex and Claude Code, adds 20+ frontier and open-weight models to both apps' model pickers, curated and kept current by the AutoRouter team. Choose a model to send requests straight to it, or choose AutoRouter and it works out how hard each task is: your plain-English rules go first, then it considers only models that clear the quality bar for the task. Hard work goes to the strongest model your plans cover, and quick work to the one that costs you least:

The landing page's examples show what that looks like (real routes follow your plans and rules): a commit message is quick work, so a low-cost open model writes it and the plan isn't touched, while a test that only fails on CI is hard, so it goes to GPT-6 Astra on the ChatGPT plan. Every request shows its model, what it cost and why. How AutoRouter picks a model walks through each step.

  1. Download AutoRouter for macOS from autorouter.com.
  2. Open it, sign in or create a free account, sign in to your plans, allow Accessibility, and click Enable for each app.
  3. Start a new session and choose AutoRouter as the model: in Codex, from the model menu under the message box; in Claude Code, with /model.

Connect Claude, ChatGPT, Grok and Gemini plans, and any of them can run in either app, so your Claude plan works inside Codex and your ChatGPT plan inside Claude Code (how that works). When one account reaches its limit, AutoRouter moves on to the account with the most room left before it spends any credit.

AutoRouter Free costs $0 a month and connects two subscriptions. Pro costs $50 a month with $20 of included usage each month, connects any number of plans and adds every other model at the provider's price plus 10%.

The trade-offs are real. ZenMux runs on any system, serves your own apps and servers as well as coding agents, pays out when output or latency falls short, and covers image and video models in its Studio. AutoRouter runs on macOS 14 or later, works with Codex (the app and the CLI), the Claude Code CLI, Cursor, Claude Desktop, T3 and Terminal, and doesn't serve the API traffic of apps you ship. Its requests pass through AutoRouter's servers on their way to each provider; it doesn't store transcripts, but keeps the message that opens each request, up to 4,000 characters, to review routing (AutoRouter privacy).

Which other ZenMux alternatives are there?

Four gateways cover what people use ZenMux for, each best at one job:

They differ most in who picks the model and what pays for requests:

ToolWho picks the modelRuns asWhat paysPrice
AutoRouterAutoRouter, per task, after your rulesMac app, macOS 14 or laterYour Claude, ChatGPT, Grok and Gemini plans first; Pro usage for other modelsFree $0; Pro $50/month with $20 of usage
ZenMuxYou, by model ID, or zenmux/auto by preferenceHosted APIZenMux credits or a Builder PlanTop-ups from $5; Builder $20, $100 or $200/month
OpenRouterYou, by model IDHosted APIOpenRouter creditsList price; 5.5% fee on credits (Standard)
RequestyYou, by model IDHosted APIRequesty balanceFree tier; 5% markup
PortkeyYou, by model IDHosted, or self-hostedYour provider keysDeveloper free; Production $49/month
LiteLLMYou, by model IDYour serversYour provider keysOpen source; Enterprise custom

What is an LLM router? explains the kinds of routers these tools are built on, and Model routers for Claude Code and Codex, compared lines up more of them.

Questions

How much does ZenMux cost?

Pay As You Go is a prepaid balance, topped up from $5, spent at each model's per-token rate. Builder Plans are $20, $100 or $200 a month for non-production use, and the free plan covers Studio Chat only (Pay As You Go, Subscription Plans).

Does ZenMux work with Claude Code and Codex?

Yes. Its Claude Code guide sets the base URL and a ZenMux key as environment variables in your shell profile, and its Codex guide adds a ZenMux provider to config.toml, which the CLI and the app both read (Claude Code guide, Codex guide).

Is ZenMux's zenmux/auto the same as AutoRouter?

No. zenmux/auto is a model name on ZenMux's API that picks a model for each request from a pool, by the preference you set, and bills your ZenMux balance (Model Routing). AutoRouter is a separate Mac app that picks a model for each task inside Codex and Claude Code, after your rules, and uses your plans first.

Does AutoRouter run on Windows or Linux?

No. AutoRouter runs on macOS 14 or later, and Claude Code needs Node 22.18 or later. On Windows or Linux, ZenMux, OpenRouter, Requesty, Portkey and LiteLLM all work, since each is an API you point your tools at.

Sources