API Platform · Public Beta

Access China's best AI models
through one clean API.

No Chinese phone number. No Alipay. No VPN.
Just an OpenAI-compatible endpoint that works everywhere.

Drop-in OpenAI SDK replacement Global CDN, sub-second latency 1/10 the price of GPT-4o
Get Free API Key → View Pricing
app.py
# Replace one line. That's it. from openai import OpenAI client = OpenAI( api_key="sk-xxx", base_url="https://api.tokensupernova.com/v1", ← here ) response = client.chat.completions.create( model="deepseek-chat", messages=[{"role": "user", "content": "Hello!"}], ) print(response.choices[0].message.content)
How it works

Three steps. One minute.

Step 1

Create account

Email signup. Get $1 free credit instantly. No credit card required.

Step 2

Get API key

One click in dashboard. Standard sk- format. Works with any OpenAI client.

Step 3

Change one line

Replace base_url in your code. Everything else stays the same.

Models & Pricing

Pay only for what you use.

DeepSeek-R1

Deep Reasoning (CoT)
$0.55
per 1M input tokens
~$2.19 per 1M output

Qwen-Max

Multilingual · Coming soon
Coming soon
 
Join waitlist

GLM-4

Chinese-optimized · Coming soon
Coming soon
 
Join waitlist

1/10 the cost. Same quality.

Provider Model Input ($/1M tokens) Output ($/1M tokens) OpenAI SDK
OpenAI GPT-4o $2.50 $10.00
Anthropic Claude 3.5 Sonnet $3.00 $15.00 ❌ (own SDK)
Google Gemini 1.5 Pro $1.25 $5.00 ❌ (own SDK)
TokenSupernova DeepSeek-V3 $0.14 $0.28
FAQ

You might be wondering.

What is this exactly?

TokenSupernova is an API gateway that connects you to Chinese large language models (DeepSeek, Qwen, GLM, etc.) through a standard OpenAI-compatible interface. You use the same code — we handle the routing, authentication, and billing.

Why would I use Chinese models?

They're significantly cheaper (often 10x) while offering comparable quality for most tasks. DeepSeek-V3 matches GPT-4o on many benchmarks at 1/18 the output cost.

Do I need a Chinese phone number or bank account?

No. That's the whole point. We handle the Chinese-side registration and payment, so you only need an email address.

Is it reliable for production use?

We run on AWS Singapore with Cloudflare CDN. Check our public status page for real-time uptime. Currently in public beta — we recommend testing before production deployment.

What about data privacy?

Your prompts and completions pass through our gateway to the upstream model provider. We don't store or log message content — only usage metadata (token counts, timestamps) for billing. See our full policy for details.