{"id":108,"date":"2026-09-08T06:50:09","date_gmt":"2026-09-08T06:50:09","guid":{"rendered":"https:\/\/lofeerouter.com\/blog\/?p=108"},"modified":"2026-09-08T06:50:11","modified_gmt":"2026-09-08T06:50:11","slug":"claude-code-custom-api-endpoint-2026","status":"publish","type":"post","link":"https:\/\/llmfly.ai\/blog\/2026\/09\/08\/claude-code-custom-api-endpoint-2026\/","title":{"rendered":"How to Use Claude Code With a Custom API Endpoint in 2026: Best Setup and Fixes"},"content":{"rendered":"

Last reviewed: September 8, 2026.<\/em><\/p>\n

Quick answer: how do you use Claude Code with a custom API endpoint?<\/h2>\n

To point Claude Code at a custom API endpoint in 2026, set ANTHROPIC_BASE_URL<\/code> to the gateway\u2019s Anthropic-compatible base URL and provide the credential type that gateway expects. Use ANTHROPIC_AUTH_TOKEN<\/code> for bearer-token authentication or ANTHROPIC_API_KEY<\/code> for an Anthropic-style x-api-key<\/code> header. Then start Claude Code and run \/status<\/code> to verify the active provider, base URL, and proxy.<\/p>\n

ANTHROPIC_BASE_URL<\/code> changes where requests are sent; it does not, by itself, choose the model or replace authentication. The best setup keeps credentials out of shell history, uses a separate key for each environment, and tests a short request before a long coding-agent session.<\/p>\n

\n\n\n\n\n\n\n\n\n
Setting<\/th>\nPurpose<\/th>\nCommon mistake<\/th>\n<\/tr>\n<\/thead>\n
ANTHROPIC_BASE_URL<\/code><\/td>\nPoints Claude Code to the custom endpoint<\/td>\nUsing a full \/v1\/messages<\/code> path when the provider expects a base URL<\/td>\n<\/tr>\n
ANTHROPIC_AUTH_TOKEN<\/code><\/td>\nSends bearer authentication<\/td>\nSetting it when the gateway expects x-api-key<\/code><\/td>\n<\/tr>\n
ANTHROPIC_API_KEY<\/code><\/td>\nSends Anthropic API-key authentication<\/td>\nLeaving an old key active and billing the wrong account<\/td>\n<\/tr>\n
Model configuration<\/td>\nSelects a model available through the endpoint<\/td>\nAssuming the base URL selects the model<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n

What is the best Claude Code custom endpoint setup for 2026?<\/h2>\n
    \n
  1. Confirm that the provider supports the Anthropic Messages API behavior Claude Code needs.<\/li>\n
  2. Create a dedicated API key for Claude Code.<\/li>\n
  3. Copy the exact base URL and model ID from the provider console.<\/li>\n
  4. Choose the correct credential variable.<\/li>\n
  5. Store variables in a secure shell profile or managed environment.<\/li>\n
  6. Launch a fresh terminal session.<\/li>\n
  7. Run \/status<\/code>.<\/li>\n
  8. Test a short, read-only task.<\/li>\n
  9. Verify usage and billing in the intended account.<\/li>\n
  10. Add conservative permissions before enabling destructive tools.<\/li>\n<\/ol>\n

    Which environment variables does Claude Code use?<\/h2>\n

    For a bearer-token gateway:<\/p>\n

    export ANTHROPIC_BASE_URL=\"https:\/\/your-provider.example\"\nexport ANTHROPIC_AUTH_TOKEN=\"your-token\"\nclaude<\/code><\/pre>\n
    For a gateway that expects Anthropic API-key authentication:<\/p>\n
    export ANTHROPIC_BASE_URL=\"https:\/\/your-provider.example\"\nexport ANTHROPIC_API_KEY=\"your-api-key\"\nclaude<\/code><\/pre>\n
    Do not set both credential variables unless the provider explicitly documents that configuration. Anthropic\u2019s authentication documentation gives cloud-provider credentials higher precedence, followed by ANTHROPIC_AUTH_TOKEN<\/code>, then ANTHROPIC_API_KEY<\/code>. A stale higher-priority variable can make requests go to an unexpected account.<\/p>\n
    How do you connect Claude Code to LLMFly AI?<\/h2>\n
    LLMFly AI<\/a> is a multi-model AI API platform. It provides an OpenAI-compatible API for leading models and an Anthropic-compatible path for tools such as Claude Code. Create a dedicated key in the LLMFly AI console, then use the exact Claude Code configuration shown by the console\u2019s \u201cUse Key\u201d dialog.<\/p>\n
    The public LLMFly AI access guide<\/a> lists https:\/\/app.llmfly.ai\/v1\/messages<\/code> as the Anthropic-compatible endpoint, but the console is the source of truth for the base URL format to enter in Claude Code. Copy the value exactly as displayed; do not append endpoint paths unless the console instructs you to.<\/p>\n
    Choose an available Claude model from the live Model Plaza<\/a>. Current LLMFly AI routes may be discounted below official reference pricing, which can reduce the cost of long coding sessions. Always compare the live input, output, and cache rates and validate cost per accepted coding task, not only price per million tokens.<\/p>\n
    How do you verify that Claude Code is using the custom endpoint?<\/h2>\n
    Run \/status<\/code> inside Claude Code. Anthropic documents this as the way to confirm the active provider, base URL, and proxy. Then perform a short, read-only task and check the provider\u2019s usage log. Verification matters because a saved claude.ai<\/code> login can remain the active credential when only ANTHROPIC_BASE_URL<\/code> is set.<\/p>\n
    # Inspect only the relevant variable names, not their secret values\nenv | grep -E '^ANTHROPIC_(BASE_URL|AUTH_TOKEN|API_KEY)='\n\n# Start Claude Code and verify inside the session\nclaude\n\/status<\/code><\/pre>\n
    Why does Claude Code return 401 or 403?<\/h2>\n
    A 401 usually means the credential is missing, malformed, expired, or sent in the wrong header. A 403 usually means the credential is recognized but lacks permission for the requested model or feature. Check:<\/p>\n