𝐗𝐀𝐈 𝐗𝐀𝐏𝐈

Configuration Guide

Unified setup notes for Codex CLI / Codex App / Claude Code / OpenCode / OpenClaw.

Codex CLI / App Claude Code OpenCode OpenClaw

Unified prerequisite: use gateway (OpenAI uses ); and configure XAI_API_KEY.

Codex CLI / Codex App

Codex supports two wire protocols: responses and chat. Choose based on model capability. Once ~/.codex/config.toml and XAI_API_KEY are configured, Codex App can reuse the same gateway setup. On Windows, the config path is %USERPROFILE%\.codex\config.toml.

Option A: wire_api = "responses" (recommended for Codex models)

# ~/.codex/config.toml
model_provider = "xai"
model = "gpt-5.4"
model_reasoning_effort = "xhigh"
plan_mode_reasoning_effort = "xhigh"
model_reasoning_summary = "detailed"
model_verbosity = "high"
approval_policy = "never"
sandbox_mode = "danger-full-access"

[model_providers.xai]
name = "xai"
base_url = ""
wire_api = "responses"
requires_openai_auth = false
env_key = "XAI_API_KEY"

# launch
export XAI_API_KEY="sk-Xvs..."
codex --yolo

Option B: wire_api = "chat" (for OpenAI Chat-shape models)

# ~/.codex/config.toml
[model_providers.xai]
name = "xai"
base_url = ""
env_key = "XAI_API_KEY"
wire_api = "chat"
requires_openai_auth = false

[profiles.minimax]
model = "MiniMax-M2.5"
model_provider = "xai"

# launch
export XAI_API_KEY="sk-Xvs..."
codex --profile minimax

Windows CMD (launch and verify)

:: Config file path
:: %USERPROFILE%\.codex\config.toml

:: Set API key
set XAI_API_KEY=sk-Xvs...

:: Option A (responses)
codex

:: Option B (chat)
codex --profile minimax

Windows PowerShell (launch and verify)

# Config file path
# $env:USERPROFILE\.codex\config.toml

$env:XAI_API_KEY="sk-Xvs..."

# Option A (responses)
codex

# Option B (chat)
codex --profile minimax

Verify with: codex (responses) or codex --profile minimax (chat)

Claude Code (gpt-5.4)

Claude Code integration is primarily environment-variable based. The following examples map Claude defaults to gpt-5.4 using .

Environment variables (Linux / macOS)

export XAI_API_KEY="sk-Xvs..."
export ANTHROPIC_AUTH_TOKEN="$XAI_API_KEY"
export ANTHROPIC_BASE_URL=""
# Optional: custom default model mapping for Claude families (not required)
export ANTHROPIC_DEFAULT_OPUS_MODEL="gpt-5.4"
export ANTHROPIC_DEFAULT_SONNET_MODEL="gpt-5.4"
export ANTHROPIC_DEFAULT_HAIKU_MODEL="gpt-5.4"

# launch Claude Code
claude

Environment variables (Windows CMD)

set XAI_API_KEY=sk-Xvs...
set ANTHROPIC_AUTH_TOKEN=%XAI_API_KEY%
set ANTHROPIC_BASE_URL=
:: Optional: custom default model mapping for Claude families (not required)
set ANTHROPIC_DEFAULT_OPUS_MODEL=gpt-5.4
set ANTHROPIC_DEFAULT_SONNET_MODEL=gpt-5.4
set ANTHROPIC_DEFAULT_HAIKU_MODEL=gpt-5.4

claude

Environment variables (Windows PowerShell)

$env:XAI_API_KEY="sk-Xvs..."
$env:ANTHROPIC_AUTH_TOKEN=$env:XAI_API_KEY
$env:ANTHROPIC_BASE_URL=""
# Optional: custom default model mapping for Claude families (not required)
$env:ANTHROPIC_DEFAULT_OPUS_MODEL="gpt-5.4"
$env:ANTHROPIC_DEFAULT_SONNET_MODEL="gpt-5.4"
$env:ANTHROPIC_DEFAULT_HAIKU_MODEL="gpt-5.4"

claude

Verify with: claude

OpenCode (Responses: gpt-5.4 / Chat: MiniMax-M2.5)

OpenCode should use the global config file ~/.config/opencode/opencode.jsonc (Windows: %USERPROFILE%\.config\opencode\opencode.jsonc). Since your system supports protocol compatibility, use one of the two complete config profiles below.

Selection order: choose API profile first (A = Responses, B = Chat), then copy the matching OS block (Linux / macOS or Windows).

Profile A (Responses API) config file (Linux / macOS)

# ~/.config/opencode/opencode.jsonc
{
  "$schema": "https://opencode.ai/config.json",
  "model": "openai/gpt-5.4",
  "small_model": "openai/gpt-5.4",
  "provider": {
    "openai": {
      "options": {
        "baseURL": "",
        "apiKey": "{env:XAI_API_KEY}"
      }
    }
  }
}

export XAI_API_KEY="sk-Xvs..."
opencode debug config

Profile A (Responses API) config file (Windows)

:: File path: %USERPROFILE%\.config\opencode\opencode.jsonc
{
  "$schema": "https://opencode.ai/config.json",
  "model": "openai/gpt-5.4",
  "small_model": "openai/gpt-5.4",
  "provider": {
    "openai": {
      "options": {
        "baseURL": "",
        "apiKey": "{env:XAI_API_KEY}"
      }
    }
  }
}

:: CMD
set XAI_API_KEY=sk-Xvs...
opencode debug config

:: PowerShell
$env:XAI_API_KEY="sk-Xvs..."
opencode debug config

Profile B (Chat API) config file (Linux / macOS)

# ~/.config/opencode/opencode.jsonc
{
  "$schema": "https://opencode.ai/config.json",
  "model": "xai-chat/MiniMax-M2.5",
  "small_model": "xai-chat/MiniMax-M2.5",
  "provider": {
    "xai-chat": {
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "",
        "apiKey": "{env:XAI_API_KEY}"
      },
      "models": {
        "MiniMax-M2.5": {}
      }
    }
  }
}

export XAI_API_KEY="sk-Xvs..."
opencode debug config

Profile B (Chat API) config file (Windows)

:: File path: %USERPROFILE%\.config\opencode\opencode.jsonc
{
  "$schema": "https://opencode.ai/config.json",
  "model": "xai-chat/MiniMax-M2.5",
  "small_model": "xai-chat/MiniMax-M2.5",
  "provider": {
    "xai-chat": {
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "",
        "apiKey": "{env:XAI_API_KEY}"
      },
      "models": {
        "MiniMax-M2.5": {}
      }
    }
  }
}

:: CMD
set XAI_API_KEY=sk-Xvs...
opencode debug config

:: PowerShell
$env:XAI_API_KEY="sk-Xvs..."
opencode debug config

Request example A: Responses API (gpt-5.4)

curl /responses \
  -H "Authorization: Bearer ${XAI_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "model":"gpt-5.4",
    "input":"Explain the purpose of a microservice gateway in one sentence"
  }'

Request example B: Chat API (MiniMax-M2.5)

curl /chat/completions \
  -H "Authorization: Bearer ${XAI_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "model":"MiniMax-M2.5",
    "messages":[{"role":"user","content":"Explain the purpose of a microservice gateway in one sentence"}]
  }'

Verify with: opencode debug config (config) and opencode run "hello" (request)

OpenClaw

OpenClaw can connect to OpenAI API and Claude API, and can also be extended to OpenAI Responses API. XAI Router supports OpenAI API and Claude API by default; for Responses mode, set api = "openai-responses". Config path: ~/.openclaw/openclaw.json on Linux / macOS, and %USERPROFILE%\.openclaw\openclaw.json on Windows.

Mode 1: OpenAI Chat API compatible (api = "openai-completions")

{
  "agents": {
    "defaults": {
      "model": { "primary": "xairouter/MiniMax-M2.5" }
    }
  },
  "models": {
    "mode": "merge",
    "providers": {
      "xairouter": {
        "baseUrl": "",
        "apiKey": "${XAI_API_KEY}",
        "api": "openai-completions",
        "models": [{ "id": "MiniMax-M2.5", "name": "MiniMax-M2.5" }]
      }
    }
  }
}

Mode 2: Claude API compatible (api = "anthropic-messages")

{
  "agents": {
    "defaults": {
      "model": { "primary": "xairouter/MiniMax-M2.5" }
    }
  },
  "models": {
    "mode": "merge",
    "providers": {
      "xairouter": {
        "baseUrl": "",
        "apiKey": "${XAI_API_KEY}",
        "api": "anthropic-messages",
        "models": [{ "id": "MiniMax-M2.5", "name": "MiniMax-M2.5" }]
      }
    }
  }
}

Mode 3: OpenAI Responses API compatible (api = "openai-responses")

{
  "agents": {
    "defaults": {
      "model": { "primary": "xairouter/gpt-5.4" }
    }
  },
  "models": {
    "mode": "merge",
    "providers": {
      "xairouter": {
        "baseUrl": "",
        "apiKey": "${XAI_API_KEY}",
        "api": "openai-responses",
        "models": [{ "id": "gpt-5.4", "name": "gpt-5.4" }]
      }
    }
  }
}

Windows CMD (env and config path)

set XAI_API_KEY=sk-Xvs...
mkdir "%USERPROFILE%\.openclaw" 2>nul
notepad "%USERPROFILE%\.openclaw\openclaw.json"

openclaw models status

Windows PowerShell (env and config path)

$env:XAI_API_KEY="sk-Xvs..."
New-Item -ItemType Directory -Path "$env:USERPROFILE\.openclaw" -Force | Out-Null
notepad "$env:USERPROFILE\.openclaw\openclaw.json"

openclaw models status

Verify with: openclaw models status

𝐗𝐀𝐈 𝐗𝐀𝐏𝐈

Verification Code Sent!

Model Cost Distribution

Model Request Distribution

Model Usage Details

Daily Usage Statistics

Account Balance Query

Query Results

🌐 Supported AI Service Providers

🧭 How It Works

Single entrypoint

Smart routing

Provider fan-out

💻 SDK Examples

OpenAI SDK Example

Anthropic SDK Example

🧪 cURL Examples

OpenAI /chat/completions

Anthropic /messages

Create New Sub-account

API Call Examples

cURL

Python

JavaScript

Fund/Deduct Sub-account

API Call Examples

cURL

Python

JavaScript

Top Up Your Account

WeChat Pay

Current Order

Customer Support

Recharge History

Configuration Guide

Codex CLI / Codex App

Option A: wire_api = "responses" (recommended for Codex models)

Option B: wire_api = "chat" (for OpenAI Chat-shape models)

Windows CMD (launch and verify)

Windows PowerShell (launch and verify)

Claude Code (gpt-5.4)

Environment variables (Linux / macOS)

Environment variables (Windows CMD)

Environment variables (Windows PowerShell)

OpenCode (Responses: gpt-5.4 / Chat: MiniMax-M2.5)

Profile A (Responses API) config file (Linux / macOS)

Profile A (Responses API) config file (Windows)

Profile B (Chat API) config file (Linux / macOS)

Profile B (Chat API) config file (Windows)

Request example A: Responses API (gpt-5.4)

Request example B: Chat API (MiniMax-M2.5)

OpenClaw

Mode 1: OpenAI Chat API compatible (api = "openai-completions")

Mode 2: Claude API compatible (api = "anthropic-messages")

Mode 3: OpenAI Responses API compatible (api = "openai-responses")

Windows CMD (env and config path)

Windows PowerShell (env and config path)

View Sub-account Information

API Call Examples

cURL

Python

JavaScript

Update Sub-account Information

Basic Information

Quota Settings

Rate Limits

Model-Specific Limits

Access Control

API Call Examples

cURL

Python

JavaScript

Sub-account List

No sub-accounts yet

Delete Sub-account

API Call Examples

cURL

Python

JavaScript

Sub-account Insights