One gateway for production AI

MiMo V2.6 is now available

Build intelligence that moves at your speed.

One gateway for high-performance reasoning, multimodal creation, voice, and agentic workflows — built for developers shipping real products.

OpenAI-compatible · Built-in usage analytics · Start free

200 OK

POST/v1/chat/completions

REQUEST

curl -X POST "/v1/chat/completions" \
  -H "Authorization: Bearer sk-••••" \
  -d '{
    "model": "your-model",
    "messages": [
      { "role": "user", "content": "..." }
    ]
}'

RESPONSE

{
  "choices": [{ "message": { "content": "Chat request routed." } }]
  "usage": { "total_tokens": 27 }
}
142 MS27 TOKENSCOST $0.00081Example
STREAMSSE

OpenAI

compatible API

8+

models and growing

Streaming

tools and batch jobs

Analytics

usage and spend tracking

Available models

Models on the gateway.

One API key, eight production models — and more added regularly.

Explore all models

DeepSeek

deepseek-v4.1-flash-fast

Fastest

DeepSeek

deepseek-v4.1-flash

Reasoning

MiMo

mimo-v2.6-pro

Flagship

MiMo

mimo-v2.6-flash

Speed

GLM

glm-5.3-flash

Balanced

GLM

glm-5.3

Reasoning

Kimi

kimi-k3

Long context

MiniMax

minimax-m3

Multimodal

Model list continues to expand — check the models page for the latest.

Built for developers

One endpoint. Every capability.

Use familiar OpenAI-compatible APIs, then route across reasoning, multimodal, speech, and batch models without rewriting your stack.

  • Unified authentication and usage tracking
  • Streaming, tools, and structured output
  • Regional endpoints and batch processing
Read the documentation

quickstart.py

1from openai import OpenAI2 3client = OpenAI(4  base_url="<gateway base URL>/v1",5  api_key="$DXP_API_KEY"6)7 8response = client.chat.completions.create(9  model="mimo-v2.6-pro",10  messages=[{"role": "user",11    "content": "Plan our launch"}],12  stream=True13)14 15for chunk in response:16  print(chunk.choices[0].delta.content)

Compatibility

Works with the tools you already use.

Use DXP credits and models from coding agents, editors, and apps — OpenAI and Anthropic-compatible, with one API key.

Official tools

Claude Code

CLI and extension — point Anthropic-compatible settings at DXP.

Codex

Coding agent — set Base URL and your key, then keep working.

OpenClaw

Coding agent and gateway with full tool-calling support.

Hermes

Add a custom provider in a few steps.

Supported editors

  • VS Code
  • Cursor
  • Antigravity
  • Claude Desktop

Compatible with OpenAI and Anthropic — any tool that can set a Base URL and API key works: agents, automation, and your own SDK.

Turn your next idea into production AI.

Start free, explore the models, and scale when you are ready.