• Home
  • Search
  • AI Services
  • Models
  • Writing
  • Multimedia Creation
  • Translation & Others
Home›Developer Center›API Pricing

API Pricing

API DocumentationAPI Quick StartOpenAI ChatGPT APIAnthropic (Claude) APIGoogle Gemini APIGoogle Search APIGoogle News APIOpenRouter APIAPI Pricing
CLI & Agent Guidescc-switch macOScc-switch Windowscc-switch LinuxOpenCode macOSOpenCode WindowsOpenCode LinuxOpenClaw macOSOpenClaw WindowsOpenClaw LinuxHermes Agent macOSHermes Agent WindowsHermes Agent LinuxClaude Code CLI macOSClaude Code CLI WindowsClaude Code CLI LinuxCodex CLI macOSCodex CLI WindowsCodex CLI Linux
MYAI168 API & MCPMYAI168 Chat API GuideMYAI168 MCP CLIMYAI168 MCP Web & App
User GuidesGT Agent GuideAI Services Guide
No matching questions on this page.Search the whole site
AI Easy Pay pricing

Pricing

  • AI Easy Pay: top up from NT$3,000. Pay only for what you use; credits are valid for one year, and large top-ups are welcome.
  • AI Easy Pay Standard annual subscription: NT$7,200 per year.
  • AI Easy Pay Premium annual subscription: NT$36,000 per year.
  • AI Easy Pay Ultra annual subscription: NT$72,000 per year.
How billing works

How API billing works

Every API on this site (the OpenAI / Anthropic (Claude) / Google / OpenRouter relays, the site-native APIs and MCP) is paid with your account credits.

  • Charged only on success: after each successful call, credits are deducted automatically based on that call's actual usage. The usage unit depends on the model and feature — for example input / output tokens, number of images generated, audio or video length, or number of web searches.
  • Never charged on errors: an invalid key, invalid parameters, a model that is not on the supported list, an upstream error or a dropped connection is never charged.
  • No request is sent without credits: when your balance is insufficient, relay endpoints return HTTP 402 right away (the site-native chat endpoint replies with a message in the stream) and the model is not called.
  • Query endpoints are free: model lists, single-model lookups, count_tokens, video job status polling and balance checks are not charged.

The exact amount charged for each call is the one reported with the response (see the next question). APIs and models change frequently. If you encounter an unsupported API or model, please contact us and we will add support as soon as possible.

Checking charges and your balance

Checking what each call cost and your balance

  • Non-streaming responses: the response headers include X-Cost-Points (points charged for this call) and X-Credits-Remaining (your balance after the call).
  • Streaming responses: after the final event, the stream ends with the SSE comment lines : cost=N and : credits-remaining=M carrying the same information; official SDKs ignore comment lines, so compatibility is not affected. Google Gemini streaming does not carry these lines — use the balance check below.
  • Google Search / News APIs: the JSON response also includes cost (points charged) and credits (balance after the call).
  • Site-native chat endpoint: a charge summary is appended at the end of the reply; for MCP, the charge and balance are in result._meta.

Check your available credits at any time (free):

curl https://www.myai168.com/en/api/user/get_credits \
  -H "Authorization: Bearer YOUR_API_KEY"
{"status": "OK", "credits": "34567", "credits_date": "2026-12-31"}

credits is your available balance (the numbers above are illustrative only) and credits_date is the expiry date of your credits; after that date the available balance is treated as 0. The Developer API Key page shows your account's cumulative API usage and the credits used by each key.

Controlling spend: API key credit limits

Controlling spend: API key credit limits

When you create or edit a key on the Developer API Key page, you can set its credit limit (0 = unlimited). Once the credits used by that key reach the limit, every further call with it returns HTTP 401 with API key credit limit reached! and the model is not called; raise the limit to continue (the used amount is not reset). The limit is checked before each call, so the last call may take the total slightly over the limit.

For MCP clients connected with OAuth (such as Claude.ai), you can set an account-level credit limit on the MCP Connections page.

We recommend a separate key, each with its own limit, for every program, CLI or AI agent you run, so one misbehaving program cannot drain your account. If a key leaks, delete it on the Developer API Key page right away and switch to a new key.

API keys and related docs

API keys and related docs

Every API authenticates with your developer key: send it in the Authorization: Bearer YOUR_API_KEY header (Anthropic endpoints also accept x-api-key, Google endpoints also accept x-goog-api-key); never place keys in URL query strings.

Get your developer key at: https://www.myai168.com/en/ai/user/developer

For base URLs, authentication, streaming, response headers and error codes, see API Quick Start; for the full list of documentation pages, see the Developer Center (https://www.myai168.com/en/developer/).

‹ Previous page: OpenRouter APINext page: cc-switch macOS ›
↑
Question link copied.

Other FAQ pages

API Documentation

  • API Quick Start
  • OpenAI ChatGPT API
  • Anthropic (Claude) API
  • Google Gemini API
  • Google Search API
  • Google News API
  • OpenRouter API

CLI & Agent Guides

  • cc-switch macOS
  • cc-switch Windows
  • cc-switch Linux
  • OpenCode macOS
  • OpenCode Windows
  • OpenCode Linux
  • OpenClaw macOS
  • OpenClaw Windows
  • OpenClaw Linux
  • Hermes Agent macOS
  • Hermes Agent Windows
  • Hermes Agent Linux
  • Claude Code CLI macOS
  • Claude Code CLI Windows
  • Claude Code CLI Linux
  • Codex CLI macOS
  • Codex CLI Windows
  • Codex CLI Linux

MYAI168 API & MCP

  • MYAI168 Chat API Guide
  • MYAI168 MCP CLI
  • MYAI168 MCP Web & App

User Guides

  • GT Agent Guide
  • AI Services Guide