Keep using existing SDKs
Connect the OpenAI SDK for TypeScript or Python, the Anthropic SDK, or direct HTTP clients to the Lurus base URL.
Compatibility API
Your existing AI tools. Now with Lurus models.
Connect OpenAI and Anthropic SDKs through one API access point. Choose an approved model, copy the snippet, and start with complete or streamed responses.
The real API workspace
1 Lurus API key
3 generation protocols
Streaming or complete response
Less integration work
One access point for familiar SDKs and approved models
Keep using established client libraries while managing models, credits, and access centrally through Lurus.
One shared model catalog
Fetch available models through /v1/models. Account and team policies determine which models are visible and usable.
The response mode your product needs
All three generation protocols support both complete responses and streaming for interactive applications.
Central billing and control
Usage draws from your existing Lurus credits. API keys can be named, assigned an expiration, and revoked immediately.
Public contract
Four endpoints. Three protocols.
The gateway deliberately supports these text-generation protocols. It does not imply coverage of every OpenAI or Anthropic endpoint.
GET
/v1/models Available models
Returns the model catalog that your account and, where applicable, your team may use.
POST
/v1/responses OpenAI Responses
For applications built on the current OpenAI Responses protocol, with complete or streamed output.
POST
/v1/chat/completions OpenAI Chat Completions
For existing Chat Completions clients and direct HTTP integrations.
POST
/v1/messages Anthropic Messages
For applications using the Anthropic Messages protocol and the x-api-key header.
Quickstart
Change one base URL
Create a Lurus API key, select an available model key, and configure the client.
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: process.env.LURUS_API_KEY,
baseURL: 'https://backend.lurus.sh/v1',
});
const response = await client.responses.create({
model: '<MODEL_KEY>',
input: 'Explain this codebase architecture.',
});The settings supported by a third-party tool depend on its current version. Follow that tool’s official documentation when configuring it.
Three steps
From dashboard to first response
The API workspace guides you from key management to a ready-to-copy example.
- 01
Open API access in the dashboard
Sign in and open API access with its Quickstart, API keys, and Models sections.
- 02
Create a named key
Choose an expiration and store the full key securely because it is shown only once.
- 03
Select a model and copy the snippet
Choose a protocol, language, response mode, and approved model, then copy the generated configuration.
Control stays centralized
Privacy, cost, and access in one place
The Compatibility API uses the existing Lurus controls for data processing, credits, model access, and key management.
Source code storage
Lurus Code processes source code and request context with Zero Data Retention.
Model training
Customer data and source code are not used to train AI models.
Billing through Lurus credits
Requests are billed by model usage against your existing personal or team credit balance.
Revocation and team policies
API keys can be revoked immediately. Team administrators centrally control which models members may use.
FAQ
Common questions about the Compatibility API
Which API protocols are supported?
OpenAI Responses, OpenAI Chat Completions, and Anthropic Messages are supported, along with the model list at /v1/models.
Can I use the OpenAI or Anthropic SDK?
Yes. The dashboard Quickstart includes examples for the OpenAI SDK in TypeScript and Python, the Anthropic SDK, and cURL.
Do the endpoints support streaming?
Yes. Responses, Chat Completions, and Messages support both complete responses and streaming.
Which models are available?
The model list depends on your account and any team policies. The /v1/models endpoint and Models section in the dashboard show the currently available options.
How are API requests billed?
Usage is billed according to the selected model against your existing personal or team credit balance.
How should I protect my API key?
Store the key in a secrets manager or environment variable, never expose it in client-side code, and revoke it in the dashboard when it is no longer needed.
Ready for your first request?
Open your API access.
Create a key, choose an approved model, and copy the matching snippet into your application.