Integrations

Use with Claude Code

Route Claude Code requests through LLM Router for fallbacks, cost optimization, and observability

Overview

Claude Code is Anthropic's agentic coding tool that runs in your terminal. By routing Claude Code through LLM Router, you get automatic fallbacks, cost optimization, and unified observability — all without changing how you use Claude Code.

Why Route Claude Code Through LLM Router?

Fallbacks: If Anthropic is rate-limited or down, LLM Router automatically falls back to another provider.
Cost control: Use model: "auto" with cost-optimized routing to reduce API spend on simpler coding tasks.
Observability: Track all Claude Code requests, token usage, and costs in the Inworld Portal.
Multi-model access: Use auto to let LLM Router pick the best model, or route to OpenAI, Google, or other providers through the same Anthropic-compatible interface.

Setup

Claude Code uses the Anthropic SDK under the hood. Redirect it to LLM Router by setting environment variables.

Add these to your shell configuration file (e.g., ~/.zshrc or ~/.bashrc):

bash

export ANTHROPIC_BASE_URL=https://api.inworld.ai
export ANTHROPIC_AUTH_TOKEN=YOUR_INWORLD_API_KEY

Claude Code sends ANTHROPIC_AUTH_TOKEN as Authorization: Bearer <key>. LLM Router accepts your API key in a Bearer header as well as Basic, so no other setup is needed. Don't use ANTHROPIC_API_KEY.

Then reload your shell and run Claude Code:

bash

source ~/.zshrc
claude

All requests from Claude Code will now route through LLM Router.

Per-Session Override

For a quick test without changing your global config:

bash

ANTHROPIC_BASE_URL=https://api.inworld.ai \
ANTHROPIC_AUTH_TOKEN=YOUR_INWORLD_API_KEY \
  claude

Using Auto Routing

By default, Claude Code sends requests to a specific Claude model. With LLM Router, you can override the model to use intelligent routing:

bash

claude --model auto

This lets LLM Router select the best available model based on your routing configuration (cost, latency, or intelligence priorities).

Using a Specific Model

You can route to any model supported by LLM Router:

bash

# Use a specific Anthropic model
claude --model anthropic/claude-opus-4-6

# Use an OpenAI model through the Anthropic interface
claude --model openai/gpt-5.4

# Use a Google model
claude --model google-ai-studio/gemini-2.5-flash

Using a Custom Router

If you've created a custom router (e.g., a cost-optimized or fallback router), reference it by ID:

bash

claude --model inworld/failover-system

Comparison: Anthropic Direct vs LLM Router

Feature	Anthropic Direct	Via LLM Router
Base URL	`https://api.anthropic.com`	`https://api.inworld.ai`
Models	Claude models only	50+ models across OpenAI, Google, Anthropic, Groq, and more
Auto routing	Not available	`model: "auto"` with configurable sort criteria (cost, latency, intelligence)
Custom routers	Not available	`model: "inworld/your-router"` with CEL conditions, variants, weights
Fallbacks	None — single provider	Automatic fallbacks across providers
Observability	Anthropic Console	Inworld Portal with routing metadata, attempt history, latency breakdown
Cost optimization	Fixed pricing	Dynamic routing to cheaper models for simple tasks

Verifying the Setup

To confirm Claude Code is routing through LLM Router, check the Inworld Portal's Observability tab after making a request. You should see the request logged with routing metadata, including which model was selected and the latency breakdown.

Troubleshooting

"Unauthorized" errors

Ensure ANTHROPIC_AUTH_TOKEN is set to your Inworld API Key. You can find your API key in the Inworld Portal under Workspace Settings > Integrations.

ANTHROPIC_API_KEY is not supported: Claude Code sends it as an x-api-key header, which LLM Router does not read. If it is set, unset it and use ANTHROPIC_AUTH_TOKEN instead.

Model not found

Ensure you're using a valid model identifier:

auto — LLM Router selects the best model
anthropic/claude-opus-4-6 — Specific model with provider/model format
inworld/your-router-id — Custom router

Connection issues

Verify the base URL is https://api.inworld.ai (no trailing slash, no path suffix).

Next Steps

Anthropic Compatibility for full API reference and feature matrix.
Cost Optimizer to reduce Claude Code API costs.
Failover System for high availability.