TokenMix Research Lab · 2026-04-30

Claude Code with OpenRouter 2026: Setup, Limits, Alternatives

Last Updated: 2026-04-30
Author: TokenMix Research Lab
Data checked: 2026-04-30

Claude Code can route through OpenRouter by overriding Anthropic environment variables. The reliable path is Anthropic first-party models through OpenRouter, not random free models.

OpenRouter's Claude Code guide says Claude Code with OpenRouter is only guaranteed to work with the Anthropic first-party provider and recommends setting Anthropic 1P as the top-priority provider. Claude Code environment docs confirm ANTHROPIC_BASE_URL, ANTHROPIC_AUTH_TOKEN, and ANTHROPIC_API_KEY control custom API routing and authentication. OpenRouter's setup requires ANTHROPIC_BASE_URL="https://openrouter.ai/api", ANTHROPIC_AUTH_TOKEN="$OPENROUTER_API_KEY", and an explicitly blank ANTHROPIC_API_KEY. That works, but it is not the same as guaranteed compatibility with every OpenRouter model.

Table of Contents

• Quick Answer
• Confirmed vs Caveat
• Setup Checklist
• Shell Setup
• Project Settings Setup
• Verify The Connection
• Model Selection Rules
• OpenRouter vs TokenMix.ai vs Direct Claude
• Cost And Limit Math
• Common Failure Modes
• When To Use Each Path
• Final Recommendation
• FAQ
• Related Articles
• Sources

Quick Answer

Use these environment variables:

export OPENROUTER_API_KEY="your-openrouter-api-key"
export ANTHROPIC_BASE_URL="https://openrouter.ai/api"
export ANTHROPIC_AUTH_TOKEN="$OPENROUTER_API_KEY"
export ANTHROPIC_API_KEY=""

Then start Claude Code:

claude

This is a useful setup for team budget controls, provider failover, and OpenRouter usage visibility. If your goal is a general multi-model API gateway for apps rather than Claude Code itself, compare TokenMix.ai and other OpenRouter alternatives.

Confirmed vs Caveat

Setup Checklist

The most common mistake is using the normal OpenRouter OpenAI-compatible base URL https://openrouter.ai/api/v1. Claude Code's OpenRouter guide uses https://openrouter.ai/api.

Shell Setup

Add this to your shell profile, such as ~/.zshrc, ~/.bashrc, or ~/.config/fish/config.fish:

export OPENROUTER_API_KEY="sk-or-your-key"
export ANTHROPIC_BASE_URL="https://openrouter.ai/api"
export ANTHROPIC_AUTH_TOKEN="$OPENROUTER_API_KEY"
export ANTHROPIC_API_KEY=""

Reload the shell:

source ~/.zshrc

Start Claude Code:

cd /path/to/your/project
claude

If Claude Code was already logged in to Anthropic, run:

/logout

Then restart Claude Code so the custom OpenRouter environment takes effect.

Project Settings Setup

Claude Code environment variables can also be configured in settings. Use this only when you intentionally want a project or team-level route.

{
  "env": {
    "OPENROUTER_API_KEY": "sk-or-your-key",
    "ANTHROPIC_BASE_URL": "https://openrouter.ai/api",
    "ANTHROPIC_AUTH_TOKEN": "sk-or-your-key",
    "ANTHROPIC_API_KEY": ""
  }
}

Do not commit real API keys. Use shell-level secrets, team secret management, or CI secrets instead.

Verify The Connection

Inside Claude Code, run:

/status

You want to see:

If /status still shows the default Anthropic route, your environment variables are not being read by the Claude Code process.

Model Selection Rules

Claude Code is not a normal chat UI. It relies on tool calls, file edits, long context, and agent loops. Model compatibility matters more than raw benchmark numbers.

My rule: use OpenRouter with Claude Code only when the route is predictable. Agent tools punish random routing.

OpenRouter vs TokenMix.ai vs Direct Claude

Use OpenRouter for Claude Code routing. Use TokenMix.ai's OpenAI-compatible API for product APIs, backend agents, and multi-model application traffic.

Cost And Limit Math

Cost calculation 1: free tier capacity

OpenRouter's pricing page lists 50 requests/day for free users, and its rate-limit docs list 20 RPM for free model variants.

Claude Code can make many model calls during one task. Counting "prompts" underestimates real usage.

Cost calculation 2: model mismatch waste

For coding agents, the cheapest model is often not the cheapest workflow.

Cost calculation 3: team budget control

OpenRouter's team value is less about "free Claude Code" and more about central visibility.

Common Failure Modes

When To Use Each Path

If your goal is to run coding agents cheaply, test end-to-end task completion, not just model response quality.

Final Recommendation

Claude Code with OpenRouter is worth using when you want OpenRouter's budgets, activity dashboard, and provider failover while staying close to Anthropic first-party compatibility.

Do not sell yourself the fantasy that any free model can replace Claude Code's native Claude route. Coding agents are tool-heavy systems. Compatibility beats novelty.

FAQ

Can Claude Code use OpenRouter?

Yes. OpenRouter has an official Claude Code guide, and Claude Code supports custom Anthropic base URL and auth token environment variables.

What base URL should I use for Claude Code with OpenRouter?

Use https://openrouter.ai/api. Do not use the normal OpenAI-compatible /api/v1 base URL for this Claude Code setup.

Why must ANTHROPIC_API_KEY be blank?

Claude Code can prefer ANTHROPIC_API_KEY for API authentication. OpenRouter's guide says to explicitly blank it so requests use ANTHROPIC_AUTH_TOKEN and the OpenRouter base URL.

Can I use free OpenRouter models with Claude Code?

You can experiment, but do not treat it as reliable production setup. OpenRouter says Claude Code with OpenRouter is only guaranteed with Anthropic first-party provider, and free models have strict limits.

How do I verify Claude Code is using OpenRouter?

Run /status inside Claude Code. It should show ANTHROPIC_AUTH_TOKEN and https://openrouter.ai/api, and requests should appear in the OpenRouter activity dashboard.

Is Claude Code with OpenRouter cheaper than direct Claude?

Sometimes, but only if the route completes tasks reliably. A cheaper model that breaks tool calls or retries can cost more in developer time.

Is TokenMix.ai a Claude Code replacement?

No. TokenMix.ai is a hosted multi-model API gateway for apps and backend agents. It is not a drop-in replacement for Claude Code's CLI runtime.

Should teams use OpenRouter for Claude Code?

Teams should consider it if they need centralized OpenRouter credits, activity visibility, and budget controls. They should still test compatibility on real repositories before rollout.

Related Articles

• OpenRouter API 2026: Pricing, Models, Limits, Alternatives
• OpenRouter Alternatives: Cheaper API Routing Options
• OpenRouter vs Direct API: When A Router Is Actually Cheaper
• Claude Code Pricing Guide 2026
• Claude Code Router Guide 2026
• Anthropic OpenAI-Compatible API 2026: Claude SDK Setup Guide
• LiteLLM Alternatives 2026: AI Gateway Options Compared

Sources

• OpenRouter Claude Code integration: https://openrouter.ai/docs/guides/guides/coding-agents/claude-code-integration
• Claude Code environment variables: https://code.claude.com/docs/en/env-vars
• Claude Code settings: https://docs.anthropic.com/en/docs/claude-code/settings
• OpenRouter pricing: https://openrouter.ai/pricing
• OpenRouter API limits: https://openrouter.ai/docs/api-reference/limits/
• OpenRouter quickstart: https://openrouter.ai/docs/quickstart