Cursor
Set the OpenAI-compatible base URL and model name.
Docs
Integrate models in 5 minutes
Start with an endpoint, API key, and model name, then finish the OpenAI-compatible setup inside your tool and review the model catalog and quota guide.
OpenAI-compatible request
base_url / key / model
from openai import OpenAI
client = OpenAI(
api_key="sk-coding-key",
base_url="https://www.prysmapi.cn"
)
response = client.chat.completions.create(
model="coding-pro",
messages=[{"role": "user", "content": "Review this patch"}]
)Quickstart
Quickstart
1Get your endpoint
2Get your API key
3Choose a model
4Send the first request
Docs index
Structured entry points
QuickstartGet your endpoint, API key, and model name, then send the first request.Tool setupFind the configuration entry points for Cursor, Codex, Cline, OpenCode, and OpenClaw.Model catalogBrowse the standard and advanced model pools by Lite and Pro coverage.Quota guideReview monthly quota, RPM / TPM expectations, and overage handling.FAQAnswers for access, model selection, quota, and activation questions.
Tool setup
Tool setup
Each tool below supports an OpenAI-compatible setup. The fields usually include base_url, API key, and model.
Codex
Configure a custom provider with your endpoint and key.
Cline
Connect through custom OpenAI-compatible providers.
OpenCode
Route coding requests through one API key.
OpenClaw
Use shared model aliases for agent workflows.
Console
Quota and console
After activation, use the console to confirm API Access, Models, Usage, Logs, and Current Plan.
API Access
Review endpoint, API key, base_url, and sample model calls.
Models
Browse the standard pool, advanced pool, and recommended models by plan.
Usage
Track monthly quota, request count, usage trends, and the active plan.
Logs
Inspect integration errors, rate limits, model responses, and request traces.
FAQ
Frequently asked questions
Questions about access, models, plans, quotas, and support.
What do I need before I integrate?
You need an endpoint, API key, model name, current plan quota, and the OpenAI-compatible configuration entry in your tool.
How should I choose a model?
Use the standard pool for everyday work, then move to the Pro pool for deeper reasoning and large-repository tasks.
How do I debug failed requests?
Check base_url, API key, model, plan coverage, monthly quota, and RPM / TPM first, then review the logs.