Get Started
All systems operational · Bangkok edge

Chinese & Global AI models,
one simple API

DeepSeek, GLM, Kimi — plus GPT, Claude, Gemini — through a single OpenAI-compatible endpoint. Sign up, grab a key, and start building in under a minute.

1 One API key for everything High-speed direct routes ¥ Pay as you go
No minimum spend · No subscription · Cancel anytime
30+Models · more coming
1API key for everything
99.9%Uptime target
<50msLatency across Asia

Works with your favorite tools

Claude Code Cursor Cherry Studio OpenCode LangChain LlamaIndex Codex OpenAI SDK
Plans

Simple, transparent pricing

Top up once, use any model. Billed per token at the rate you see — no surprises.

Free Trial

Start exploring without a credit card.

  • Free models like glm-4.7-flash
  • Instant registration
  • No card required
Start Free
RECOMMENDED

Pay-as-you-go

Flexible top-up, billed by actual usage.

  • Billed per token, per model
  • Top up with international cards
  • Usage dashboard & call logs
  • Balance works across 30+ models
Top Up & Go

Enterprise

Custom capacity and dedicated support.

  • Dedicated deployment
  • 24/7 expert support
  • SLA guarantee
Contact Sales
Quick start

From zero to your first call in 3 minutes

One key, one endpoint, all models. Four steps and you're live.

01

Register

Sign up with email or OAuth — no credit card needed.

02

Copy your API key

Create a token from the dashboard and copy it.

03

Set your base URL

Point your client at https://www.mlandi.com/v1.

04

Make the call

Pick a model and send your first request. Done.

Models

Model marketplace

One key, one bill, one integration — every model below and more.

DeepSeekDirect connection

deepseek-v4-flash$0.88/M in
deepseek-v4-pro$2.64/M in

Zhipu GLMDirect connection

glm-4.7-flashFREE
glm-5 · 5.1 · 5.2$0.73–1.44/M
glm-5.3NEW

Moonshot KimiDirect connection

kimi-k3$3.61/M in
kimi-k2.7-code$1.17/M in
kimi-k2.6$1.17/M in

QwenAlibaba · Coming soon

qwen-3.7-flashSOON
qwen-3.7-maxSOON

MiniMaxComing soon

minimax-m3SOON
minimax-m2.7SOON

DoubaoByteDance · Coming soon

doubao-seed-2.1-proSOON
doubao-seed-2.1-turboSOON

Global ModelsGPT · Claude · Gemini

gpt-5.6-luna$0.27/M in
claude-sonnet-5$2.74/M in
gemini-3.5-flash-lite$0.41/M in
Prices are input rates per 1M tokens · Full list & live pricing on the pricing page
Why Mlandi

Built for teams outside China

Chinese-model performance, without the friction.

OpenAI-compatible

Drop-in replacement — change your base URL and key, keep your SDKs and agents.

🌏

Bangkok edge

Hosted in Thailand — direct routes to Chinese providers, fast from Asia and beyond.

💳

Easy payments

Top up with international cards. No Chinese bank account or Alipay needed.

🤖

24/7 AI support

Instant answers from our assistant in the corner, plus email for anything else.

Comparison

Mlandi vs. going direct

With Mlandi

  • One key & one bill for 30+ models — switch by changing one string.
  • Pay like a local — international cards work out of the box.
  • English docs & support for every model.
  • Unified rate limits, logs and dashboards.

Direct with each provider

  • Separate accounts and invoices for every provider.
  • Many Chinese providers can't accept foreign cards.
  • Docs and support mostly in Chinese only.
  • Fragmented rate limits and tracking.
FAQ

Questions, answered

Is it really compatible with OpenAI SDKs?

Yes. We expose a standard OpenAI-compatible API at /v1. Point your client at our base URL, swap in your key, and chat completions, streaming and tool calls all work as expected.

How does pricing work?

Simple pay-as-you-go: billed per token at the listed rates. No subscriptions, no minimum spend. Try free models first — glm-4.7-flash costs nothing.

Can I use GPT, Claude and Gemini here too?

Yes. Alongside DeepSeek, GLM and Kimi, we offer leading global models under the same key — ideal for A/B testing Chinese and Western models side by side.

What happens if I run out of balance?

Requests return a quota error once balance hits zero — no surprise overages. Top up anytime and you're back instantly.

Where are your servers located?

Our gateway runs in Bangkok, Thailand — low latency across Southeast Asia, with good routes to both Chinese providers and global users.

Ready to build?

Get your API key in 60 seconds. Start free, scale when you're ready.

Create Free Account
mlandai · One API for Chinese & global AI models · © 2026
support@mlandi.com