❓ Frequently Asked Questions

Go to the registration page, enter your email and password. No phone number, no ID, no real-name verification required.

On the login page, click "Forgot password?" and enter your email. We'll send you a reset link.

This is normal — sessions expire after a period of inactivity for security. Just log in again.

We accept credit/debit cards, PayPal, and cryptocurrency. All payments are processed instantly. No WeChat Pay or Alipay required.

Payments are processed instantly. If your balance hasn't updated after a successful payment, refresh the page or contact support.

Topped-up balances are non-refundable. We recommend starting with a smaller amount ($10–$20) to test the service first.

Quota Top-Up: Directly add funds to your wallet — pay as you go.
Subscription Plans: Pre-purchase a bundle with bonus credits (e.g., $100 plan gives $120 credit). Subscriptions are better for heavy users.

We are 100% OpenAI SDK compatible. Just change your base_url to https://www.tokencnn.com. We also support Claude native format (/v1/messages) and Gemini native format.

Log in to the console → Token Management → Create a new token. New accounts get a default token automatically.

Pass the model name in the model parameter of your API request. Call GET /v1/models or check the console for the full list.

Yes! Set stream: true in your request to get SSE (Server-Sent Events) streaming, identical to OpenAI's implementation.

30+ models including:

• DeepSeek — DeepSeek-V4, DeepSeek-Reasoner
• Alibaba Cloud — Qwen-Max, Qwen-Plus, Qwen-Turbo
• Zhipu AI — GLM-4, GLM-4-Flash
• Moonshot AI — Kimi-K2
• ByteDance — Doubao-Pro
• OpenAI — GPT-4o, GPT-4o-mini
• Anthropic — Claude-Sonnet, Claude-Haiku
• Google — Gemini series

Chinese AI labs (DeepSeek, Alibaba, Zhipu, etc.) produce world-class models, but accessing them directly requires Chinese phone numbers, real-name ID verification, WeChat Pay, and Chinese-language documentation. tokencnn.com removes ALL these barriers — you get the same models with an email signup and a credit card.

Log in to the console → Dashboard for your balance and usage overview. Detailed logs are in Usage Logs.

API calls will return a 402 Payment Required error. Top up to resume service immediately.

Possible causes:

• Incorrect API key — check your Authorization header
• Key has been revoked or disabled — check in Token Management
• Wrong format — use Bearer sk-xxx

Rate limit exceeded. Solutions:

• Wait 1-2 seconds and retry
• Reduce your request frequency
• Upgrade your plan for higher limits

Temporary backend issue. Retry after a few seconds. If the problem persists, contact support with the request ID from the x-oneapi-request-id response header.

Timeouts can happen due to:

• Slow model inference (especially for large reasoning models)
• Network issues
• Very long input content

Log in, click the 「≡」 menu on the top-left sidebar, then select Playground.

Pick a model and start chatting — no client configuration needed.

• ⚡ Browser-ready — just open and use.

Give it a try!