⚡ DeepSeek V4 · OpenAI Compatible · Global Edge

The fastest way to use
DeepSeek V4 from anywhere.

OpenAI-compatible endpoint. Pay with any credit card. No VPN, no Alipay, no Chinese phone number required. One baseURL change and you're shipping.

🚀

OpenAI Compatible

Drop-in replacement. Change one line — your baseURL — and your existing OpenAI SDK keeps working.

🌍

Global Edge Network

Powered by Cloudflare Workers running in 300+ cities. Low latency from any country, no proxy needed.

💳

Pay Like a Normal Person

Visa, Mastercard, Stripe, crypto. No Alipay, no Chinese ID, no SMS code from a Chinese phone number.

Quickstart

Replace OPENAI_API_KEY and OPENAI_BASE_URL. That's it.

curl
Python
Node.js
# 1. Get your key
curl https://api.tokenguo.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-dgw-yourname-xxxx" \
  -d '{
    "model": "deepseek-v4-flash",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
from openai import OpenAI

client = OpenAI(
    api_key="sk-dgw-yourname-xxxx",
    base_url="https://api.tokenguo.com/v1"
)

resp = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(resp.choices[0].message.content)
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "sk-dgw-yourname-xxxx",
  baseURL: "https://api.tokenguo.com/v1",
});

const resp = await client.chat.completions.create({
  model: "deepseek-v4-flash",
  messages: [{ role: "user", content: "Hello!" }],
});
console.log(resp.choices[0].message.content);

Pricing

Pay only for what you use. No subscriptions. No hidden fees.

Free Beta
$0
  • 100k tokens / day
  • v4-flash model only
  • Email support
  • For evaluation only
Get a Test Key
Self-Hosted
$999 one-time
  • We deploy on your Cloudflare
  • You own the infrastructure
  • No middleman after delivery
  • 30-day support included
Talk to Us

FAQ

Common questions from developers like you.

Why not call DeepSeek directly?

You can — if you have a Chinese phone, Alipay, and live next to a low-latency POP. Most international devs don't. We solve all three with one URL change.

How is this different from OpenRouter / Together / DeepInfra?

We focus on DeepSeek only and ship Cloudflare-edge latency without their margin stack. Smaller, faster, cheaper for this one use case.

Is this stable enough for production?

We're in beta. The infrastructure (Cloudflare Workers) has 99.99% SLA. Our token-counting and rate-limiting are battle-tested in beta. We'll let you know before any breaking change.

What happens to my data?

We pass your prompts straight to DeepSeek and stream the response back. We log token counts for billing. We do not store prompt or completion content.

Can I run this myself?

Yes — see the Self-Hosted tier. We give you the Cloudflare Worker code, deploy it on your account, and walk you through it. After that you own everything.

Get a free test key

Reply with your project, we'll send a key within hours.

📧 Email hello@tokenguo.com