# Astron MaaS · Astron Token Plan User Guide

# 1. Key Configuration Information (Required Reading)

The following are the core parameters required to integrate with Astron Token Plan. Copy them into the configuration file for the corresponding tool.

# 1.1 Obtain an API Key

1. Visit the Astron Token Plan subscription page on the Astron MaaS platform (opens new window).
2. After purchasing a plan, copy its dedicated Astron Token Plan API Key.

Each plan has a separate API Key that can only be used with the Astron Token Plan API.

# 1.2 API Endpoints

# 1.3 Model Names

Specify a model ID in the configuration file to switch models. The following model IDs are currently supported:

# 2. Tool Configuration Templates

# 2.1 OpenClaw

Configuration file path: ~/.openclaw/openclaw.json

{
  "models": {
    "mode": "merge",
    "providers": {
      "astrontokenplan": {
        "baseUrl": "https://maas-token-api-lac.xf-yun.com/v2",
        "apiKey": "YOUR_TOKEN_PLAN_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "YOUR_CONFIGURED_MODEL_ID",
            "name": "YOUR_CONFIGURED_MODEL_ID",
            "reasoning": false,
            "input": ["text"],
            "contextWindow": 92160,
            "maxTokens": 32768
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "astrontokenplan/YOUR_CONFIGURED_MODEL_ID"
      },
      "models": {
        "astrontokenplan/YOUR_CONFIGURED_MODEL_ID": {
          "alias": "astrontokenplan"
        }
      }
    }
  }
}

# 2.2 Claude Code

Configuration file path: ~/.claude/settings.json

{
  "env": {
    "ANTHROPIC_AUTH_TOKEN": "YOUR_TOKEN_PLAN_API_KEY",
    "ANTHROPIC_BASE_URL": "https://maas-token-api-lac.xf-yun.com/anthropic",
    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": 1,
    "API_TIMEOUT_MS": 600000,
    "ANTHROPIC_MODEL": "YOUR_CONFIGURED_MODEL_ID",
    "ANTHROPIC_SMALL_FAST_MODEL": "YOUR_CONFIGURED_MODEL_ID"
  },
  "permissions": {
    "allow": [],
    "deny": []
  }
}

Optional configuration: ~/.claude.json

{
  "hasCompletedOnboarding": true
}

# 2.3 Cursor

Configuration location: Press Ctrl + Shift + J, then search for “Models”

# 2.4 OpenCode

Configuration file path: opencode.config.json in the project root, or the configuration path specified by OpenCode (refer to the OpenCode documentation (opens new window))

Astron Token Plan uses a dedicated Base URL (maas-token-api), which differs from the standard inference service (maas-api). Do not mix them.

Complete configuration example:

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "AstroTokenPlan": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Astron Token Plan",
      "options": {
        "baseURL": "https://maas-token-api-lac.xf-yun.com/v2",
        "apiKey": "YOUR_TOKEN_PLAN_API_KEY"
      },
      "models": {
        "astron-code-latest": {
          "name": "YOUR_CONFIGURED_MODEL_ID"
        }
      }
    }
  }
}

# 3. Product Overview

Astron Token Plan on the Astron MaaS platform is a monthly subscription service that gives businesses and teams access to large language models, with usage measured uniformly in credits. Allowances are allocated per member, and each member receives a fixed amount. Within that allowance, users can access multiple large language models, with different models consuming different amounts of credits. The service supports both the OpenAI and Anthropic protocols. It is compatible with mainstream AI coding tools such as OpenClaw, Cursor, and Claude Code, covering scenarios including code writing, comprehension, and optimization to help improve development efficiency and code quality.

# 4. Plans and Billing

# 4.1 Plan Allowance Consumption

Credits are deducted for each request based on actual model usage. You can view credit usage on the Astron Token Plan subscription page on the Astron MaaS platform (opens new window).

Credit consumption follows a unified calculation rule. Refer to the model list for the amount consumed by each model.

Credits consumed = Non-cached input tokens (millions) × Input credits
                   + Cache-hit tokens (millions) × Cache-hit credits
                   + Output tokens (millions) × Output credits
                   + Reasoning tokens (millions) × Reasoning credits (if reasoning mode is available)

Actual credits consumed = Credits consumed × Time-period factor

# 5. Supported Models

# 6. Subscription Management

# 6.1 Subscription and Purchase

1. Sign in to the Astron MaaS platform and open the Astron Token Plan Subscription page.
2. Select Standard Member, Advanced Member, or Premium Member. After payment for a plan subscription is completed, the plan's effective period and rate-limit cycle begin. Each monthly plan is valid for 31 days from the completion of payment.
3. On the plan subscription page, you can view the following information (subject to the content actually displayed on the page): the current plan, periodic usage, the API Key, and the upgrade option.

# 6.2 Plan Stacking and Downgrades

• Stacking: You can make multiple purchases at the same tier or mix purchases across different tiers.
• Downgrade: You cannot downgrade a currently active Premium Member or Advanced Member plan. To use a lower-tier plan, wait until the currently active plan expires, then purchase the lower-tier plan again.

# 6.3 Cancellation and Refunds

Once purchased, the Astron Token Plan service cannot be canceled or refunded except in the following circumstances. Please place your order carefully: (1) the user cannot access the service after payment due to a platform issue; or (2) a refund is otherwise required by laws or regulations.

# 6.4 Effective Time

• Plan models: Specify a modelId in the configuration file to switch models easily.

# 7. Agreements and Policies

• This service is governed by the Astron MaaS Platform User Agreement and relevant product terms.
• By using Astron Token Plan, you agree to the allowance purposes and usage restrictions stated in this product documentation.
• Refer to the console and the latest announcements for the applicable SLA and detailed billing rules.

# 8. Error Codes

# 9. Contact Us

If you have any questions, encounter an issue, or need technical support, please contact our support team using the email address below. We will be happy to assist you.

Email address: support@astron.ai (Send email)