OpenAI-, Claude- & Gemini-compatible

One API for every AI model.

Access GPT, Claude, Gemini, DeepSeek and dozens more through a single endpoint. Better prices, higher uptime, low latency.

Drop-in compatible — your OpenAI SDK just works
request.sh
# Point your OpenAI client at VietToken
curl https://api.viettoken.app/v1/chat/completions \
  -H "Authorization: Bearer $VIETTOKEN_KEY" \
  -d '{
    "model": "openai/gpt-4o-mini",
    "stream": true,
    "messages": [{"role":"user","content":"Hi!"}]
  }'
▎stream
89+
Models
35+
Providers
~50ms
Gateway overhead
99.9%
Uptime target
Works with the tools you already use
OpenAI SDKAnthropicGeminiLangChainVercel AI SDKCursorn8n
How it works

One request, every model

Send one request to a single endpoint. VietToken routes it to the best model and streams tokens straight back — with automatic failover.

💻
Your app
openai SDK
VietToken
gateway · 1 endpoint
AI
openai/gpt-5.x
frontier
AN
anthropic/claude-opus-4
reasoning
GE
google/gemini-3-pro
multimodal
DS
deepseek/deepseek-v4
reasoning
XA
x-ai/grok-4
frontier
QW
qwen/qwen-3.5
open
KI
moonshot/kimi-k2.5
long-context
MI
mistral/mistral-large
fast
Auto-failover: if a model or key goes down, traffic reroutes instantly — your app never breaks.
Why VietToken

What you actually need

One integration, every model. Fast, reliable, and easy to plug in.

Low latency

Edge servers + unbuffered streaming for near-instant first tokens. ~30–90ms overhead.

🔀

Any model, one API

OpenAI, Anthropic, Google, DeepSeek, xAI and 35+ providers — switch with one line.

🛡️

Higher uptime

Automatic failover across keys and providers. When one fails, traffic reroutes instantly.

🔌

Drop-in compatible

Keep your existing code. Just change the base URL — no rewrites, no lock-in.

Get started

Live in three steps

From sign-up to your first streamed token in under a minute.

1

Create an account

Sign up free and get trial credit.

2

Add credits

Top up once, spend on any model. No subscriptions.

3

Get your API key

Create a key and start making requests.

base_url = "https://api.viettoken.app/v1"
One key — every model

The latest models, instantly

Access frontier models through one unified API. Add or remove anytime.

GPT-5.x GPT-4o mini Claude Opus 4 Claude Haiku 4.5 Gemini 3 Pro Gemini 3 Flash DeepSeek V4 Qwen 3.5 Grok 4 Kimi K2.5 Mistral Large + your custom model
Pricing

Pay as you go. No surprises.

Buy credits once, use on any model. No subscriptions, no hidden fees.

Free

For trials & side projects
$0
  • $1 trial credit
  • Access to all models
  • 3 API keys
  • Community support
Start free
Most popular

Pro

For growing products
At cost
  • No per-token markup — pay model price
  • Unlimited tokens & keys
  • Private custom providers
  • Load-balancing & failover
  • Priority support
Add credits

Enterprise

For teams & scale
Custom
  • Volume-based discounts
  • SLA & uptime guarantees
  • SSO, roles & data policies
  • Dedicated 24/7 support
Contact sales
Quickstart

Integrate in 60 seconds

OpenAI-compatible — just change the base URL.

request.sh
curl https://api.viettoken.app/v1/chat/completions \
  -H "Authorization: Bearer $VIETTOKEN_KEY" \
  -d '{ "model":"openai/gpt-4o-mini", "messages":[{"role":"user","content":"Hello"}] }'
from openai import OpenAI

client = OpenAI(
    base_url="https://api.viettoken.app/v1",
    api_key="$VIETTOKEN_KEY",
)
stream = client.chat.completions.create(
    model="anthropic/claude-haiku-4.5",
    messages=[{"role":"user","content":"Hello"}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.viettoken.app/v1",
  apiKey: process.env.VIETTOKEN_KEY,
});
const stream = await client.chat.completions.create({
  model: "google/gemini-3-flash",
  messages: [{ role: "user", content: "Hello" }],
  stream: true,
});
FAQ

Questions, answered

Is it really OpenAI-compatible?+

Yes. Use the official OpenAI SDKs — just set the base URL to VietToken and your existing code works unchanged.

Which providers are supported?+

OpenAI, Anthropic, Google, DeepSeek, xAI, Mistral, OpenRouter and 35+ types, plus any OpenAI-compatible custom endpoint — including self-hosted vLLM/Ollama.

How do you keep it fast?+

Edge servers close to your users plus pass-through SSE streaming with no buffering. The gateway adds only ~30–90ms.

How does billing work?+

Pay as you go: buy credits once and spend them on any model. Track usage, latency and cost per key. No subscriptions.

Start building with every model today

Get a free API key, integrate in minutes, pay only for what you use.

Get your free API key →