OpenAI-compatible API gateway

Lowertoken

Use the latest Chinese AI models through one OpenAI-compatible API. Route to DeepSeek V4, Kimi K3, GLM-5.3, Qwen 3.6, MiniMax M3, MiMo V2.5 and more while keeping one key, one bill and transparent model IDs.

Create API key View launch pricing Read data flow

Drop-in OpenAI SDK / Early access accounts / Creem checkout in review / BTCPay crypto planned / No prompt training by Lowertoken

China model route board - live catalog
Balance$25.00
Today1.8M
Saved38%
deepseek-v4-proFlagship reasoning and coding
Mid
Popular
Online
kimi-k3Latest Kimi general and agent model
Low
New
Online
glm-5.3Reasoning and agent workflows
Mid
New
Online
qwen3.6-plusGeneral chat and agent workflows
Mid
New
Online
client = OpenAI(
    api_key="LOWERTOKEN_API_KEY",
    base_url="https://api.lowertoken.com/v1"
)
1 keyOpenAI-compatible chat completions endpoint
49 liveEnabled model routes verified in the production catalog
$10+Small prepaid top-ups for early builders

Built for cost-sensitive builders

Route efficient models without rebuilding your stack.

Lowertoken sits between your app and selected upstream providers. It handles account keys, prepaid balance, routing notes and usage visibility while your code keeps the OpenAI client shape.

01

Current China model menu

Choose current flagship, fast and coding routes across DeepSeek, Kimi, GLM, Qwen, MiniMax, MiMo, Doubao, Step and Hunyuan families.

02

Transparent data flow

Every customer-facing page points to the same routing policy: gateway checks, selected upstream, provider processing and metadata retention.

03

Prepaid usage control

Prepaid balances, per-key limits and visible usage make it practical for products, agents, evaluation workloads and team projects.

Updated model families

The latest Chinese models, with exact IDs developers can use.

See route status
DeepSeek Kimi MiniMax Qwen GLM MiMo Doubao Step Hunyuan

Data flow first

Your prompt route should not be a black box.

For enterprise users in Japan, Korea, Europe and North America, the first question is where prompts go. Lowertoken makes that route visible before you send production traffic.

Your appOpenAI-compatible request with your Lowertoken key.
LowertokenAuth, balance, rate limit and route selection.
Upstream providerSelected model endpoint processes the prompt under its own terms.
ResponseStreaming response returns through the gateway to your client.

Two-line migration

Use the SDK you already have.

Open quickstart 
from openai import OpenAI

client = OpenAI(
    api_key="LOWERTOKEN_API_KEY",
    base_url="https://api.lowertoken.com/v1",
)

response = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[{"role": "user", "content": "Ship a concise launch plan."}],
)