⚡ DeepSeek V4 · 90% cheaper than GPT-4o

Access China's Best
AI Models. One API Key.

No Chinese phone number required. No ID verification. Just a single API key that works with OpenAI's SDK — plug and play.

Simple, Transparent Pricing

Pay only for what you use. No monthly fees. No minimum commit.

⚡ Best Value

DeepSeek V4-Flash

deepseek-v4-flash
Input (per 1M tokens)$0.56
Output (per 1M tokens)$1.12
Context Window128K
🚀 Pro

DeepSeek V4-Pro

deepseek-v4-pro
Input (per 1M tokens)$2.80
Output (per 1M tokens)$5.60
Context Window128K
🧠 Reasoning

DeepSeek R1

deepseek-reasoner
Input (per 1M tokens)$2.20
Output (per 1M tokens)$8.76
Reasoning✓ Included
💡 New users get $1.00 free credit — that's ~1.8M tokens with V4-Flash. No credit card required to start.

How we compare

DeepSeek V4-Flash
$0.56/M input
GPT-4o: $2.50/M
DeepSeek V4-Pro
$2.80/M input
Claude Sonnet 4: $3.00/M
DeepSeek R1
$2.20/M input
o1: $15.00/M

Why Token12?

📱

No Chinese Phone Required

Sign up with just your email. No +86 phone verification, no real-name ID checks. You're in, in 30 seconds.

🔌

OpenAI-Compatible API

Drop-in replacement. Just change the base URL and API key. Works with OpenAI SDK, LangChain, LlamaIndex, and every major framework.

💳

Pay with Card / Crypto

Pay with Visa, Mastercard, or USDT. Auto top-up. No minimum deposit. Pay only for what you use.

Quick Start

Works with any OpenAI SDK. One line change.

1

Get your API key

Sign up at token.12fz.com
# $1 free credit on registration
2

Set your base URL

# Python with OpenAI SDK
client = OpenAI(
api_key="sk-your-key",
base_url="https://token.12fz.com/v1"
)
3

Make your first call

response = client.chat.completions.create(
model="deepseek-v4-flash",
messages=[{"role": "user", "content": "Hello!"}]
)

FAQ

How do I get an API key?
Sign up at token.12fz.com with just your email. You'll receive $1 free credit immediately. No phone number or ID needed.
What payment methods do you accept?
Currently we accept USDT (ERC-20 / TRC-20). Send USDT to our wallet address, and your account will be credited within minutes. Credit/debit card via Stripe coming soon.

0xc3ce2262317e80eDfeBeE58e6004A1Cbd944E63A Send USDT (ERC-20 or TRC-20) and email your tx hash to get credited. Minimum top-up: 10 USDT (~$10).
Can I use this with LangChain / LlamaIndex / Cursor?
Yes! Any tool that supports OpenAI-compatible APIs. Just set the base URL to https://token.12fz.com/v1 and use your API key.
What's the latency like from outside Asia?
Our relay runs on Vultr Tokyo (Japan). Typical latency: ~50ms (Asia), ~120ms (US West), ~180ms (Europe). Fast enough for most applications.
Is there a rate limit?
Default rate limit is 60 RPM for new accounts. Higher limits available on paid plans. Contact us if you need higher throughput.