​
What is OpenResponses

A fully OpenAI-compatible, self-hostable API — enabling developers and enterprises to build and deploy AI applications faster while retaining full control of data, models, and infrastructure. If it works with OpenAI — agents, SDKs, client libraries, and tools — it works with OpenResponses, without changes. Self-host to maintain security, compliance, and flexibility in choosing any model while leveraging the full speed and evolving capabilities of the OpenAI ecosystem.

​
Supported Tools

  • OpenAI Agent SDK, OpenAI Clients, Any framework (LangChain, LlamaIndex, etc.) that works with OpenAI’s Responses API.

​
Deliver AI apps faster without Lock-in

  • Enterprise-Grade Data Privacy: Keep sensitive information fully under your control—no external dependencies on model provider platforms.
  • All-in-One Platform: Eliminate the hassle of juggling multiple disjointed services. Access powerful builtin tools that is extendbale through a single API.
  • Instant ROI: Shorten time-to-value with prebuilt integrations and automated tracing for auditing performance and usage.
  • Compliance Assured: Meet regulatory requirements and internal security policies with a fully self-managed solution.

​
Empower AI Agent Builders

​
Focus on Innovation, Not Infrastructure

OpenResponses supercharges your ability to build AI-driven agents and apps by providing the essential tools out of the box.
  • Automatic Data Retrieval (RAG): Enhance chatbot or agent responses with real-time access to external and internal data.
  • Integrated Tooling: Leverage built-in web search, GitHub access, and more—no tedious setup required.
  • Unified API Interface: Add advanced AI features to your apps simply by calling an OpenAI-compatible endpoint.
  • Real-Time Monitoring: Gain immediate insights into performance, usage patterns, and debug logs through automated tracing.

​
Get Started in Minutes

​
Launch with a Single Docker Command

Whether you’re an experienced dev or just exploring AI, OpenResponses makes it dead simple to spin up your own AI environment. 
docker run -p 8080:8080 masaicai/open-responses:latest

​
Seamless Integration

If you already use the OpenAI SDK, just point your API calls to OpenResponses. Minimal code changes—maximum productivity. 
from openai import OpenAI
import os

openai_client = OpenAI(
    base_url="http://localhost:6644/v1",
    api_key=os.getenv("XAI_API_KEY")
)

response = openai_client.responses.create(
    model="xai@grok-4-0709",
    input="Tell me a joke"
)

​
Using with OpenAI Agent SDK

from openai import AsyncOpenAI
from openai.agents import Agent, OpenAIResponsesModel
import os

client = AsyncOpenAI(
    base_url="http://localhost:6644/v1",
    api_key=os.getenv("XAI_API_KEY")
)

agent = Agent(
    name="Assistant",
    instructions="You are a humorous poet who can write funny poems of 4 lines.",
    model=OpenAIResponsesModel(model="xai@grok-4-0709", openai_client=client)
)

​
Using with cURL

curl --location 'http://localhost:6644/v1/responses' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer XAI_API_KEY' \
--data '{
    "model": "xai@grok-4-0709",
    "stream": false,
    "input": [
        {
            "role": "user",
            "content": "Tell me a joke"
        }
    ]
}'

​
Build and Iterate Faster

Focus on creating AI-driven experiences, not wrestling with infrastructure. OpenResponses gives you a production-ready, enterprise-capable AI toolkit from day one.

​
Questions?

Check out our Frequently Asked Questions for answers to common questions about OpenResponses, or proceed to the Quick Start Guide to begin using OpenResponses right away.

​
Future Development

Interested in where Platform is headed? View our detailed Development Roadmap to learn about upcoming features and enhancements.