Routing
One endpoint for both model markets
Use the same OpenAI-compatible request shape across China AI models and global providers. Switch models with the model parameter.
OpenAI-Compatible Multi-Model API
One API for global and China AI models.
Use one base_url to access DeepSeek, Qwen, Kimi, GLM, OpenAI, Claude, Gemini, and more. AITransX gives developers unified API keys, model routing, usage billing, request logs, and failover for practical multi-model apps.
OpenAI SDK compatible
Request and cost logs
Global + China model routing
1
base_url for your SDK
China
DeepSeek, Qwen, Kimi, GLM
Global
OpenAI, Claude, Gemini
Logs
request and cost visibility
Integration
Change the base_url. Keep your OpenAI SDK.
Start by replacing the SDK base_url, then choose a model ID such as deepseek-chat, qwen-max, kimi-k2, gpt-4o, or claude-3.5-sonnet.
deepseek-chat
qwen-max
kimi-k2
glm-4.5
gpt-4o
claude-3.5-sonnet
gemini-flash
Python
from openai import OpenAI client = OpenAI( api_key="aitx-your-key", base_url="https://aitransx.com/v1", ) response = client.chat.completions.create( model="deepseek-chat", messages=[{"role": "user", "content": "Hello"}], )
Product
A simple gateway with the controls teams actually need.
Keep the SDK integration small while keeping model access, usage, cost, and privacy visible.
Reliability
Failover without SDK rewrites
When an upstream route fails, AITransX can try the next healthy provider while your application keeps the same API surface.
Billing
Usage you can inspect
Track requests, latency, token usage, and cost by key, model, and provider. Export logs when your team needs an audit trail.
Security
Prompt bodies are not stored
Request metadata is logged for operations; prompts and responses are streamed through without being written to disk.
Why teams use it
Built for practical multi-model API work.
Clear usage-based spend
Start with free credit, then pay by usage. Dashboard balance and per-model spend make costs easier to understand.
Useful request visibility
See request status, latency, provider, model, token usage, cost, and request IDs from the dashboard.
China and global model coverage
Use DeepSeek, Qwen, Kimi, and GLM alongside OpenAI, Claude, Gemini, and other global models through one API.
Use cases
Useful when your users, models, and infrastructure are not all in one market.
Teams outside China using China AI models
Access DeepSeek API, Qwen API, Kimi API, GLM API and other China AI model routes through an OpenAI-compatible gateway.
China-based teams using global models
Call OpenAI, Claude, Gemini, Mistral, Grok-style and other global model APIs through one familiar interface.
Multi-model products with one control surface
Use different models for different products or regions while keeping API keys, billing, request logs, and routing in one place.
Create a key and send your first request.
No vendor migration project. No model-by-model SDK rewrite. Start with the OpenAI SDK pattern you already know.