Compatibility Matrix

CheapAI exposes an OpenAI-compatible API. The matrix below lists integrations that have been tested by the CheapAI team, with dates and proof snippets. Untested rows are marked accordingly.

Testing Methodology

Tested = CheapAI team reproduced a successful chat completion via the integration on the listed date. Evidence is a working code snippet or configuration block.
Partially tested = Basic chat completions work; streaming or advanced features (function calling, vision) not yet systematically validated.
Not yet tested = Integration has not been formally validated by the CheapAI team. May still work given the OpenAI-compatible protocol.

Testing was performed against the CheapAI proxy endpoint using real API keys. All tests used the claude-sonnet-4-6-20260217 or gpt-5.4 model IDs unless noted.

Key: Tested Partially Tested Not Yet Tested

Integration / SDK	Status	Scope & Tested Date	Proof Snippet
Official OpenAI Node.js SDK openai@latest (npm)	Tested	Chat completions (streaming & non-streaming) Tested: March 2026	`const openai = new OpenAI({ baseURL: 'https://cheapai-netifly-app.up.railway.app/v1', apiKey: 'YOUR_CHEAPAI_KEY' });`
Official OpenAI Python SDK openai>=1.0.0 (pip)	Tested	Chat completions, streaming, system prompts Tested: March 2026	`from openai import OpenAI client = OpenAI( base_url="https://cheapai-netifly-app.up.railway.app/v1", api_key="YOUR_CHEAPAI_KEY" )`
Cursor AI IDE Custom API key mode	Tested	Code completion, chat mode, Composer Tested: March 2026 — see our Cursor setup guide	Settings → AI → Custom API: Set Base URL + key. Model: `claude-sonnet-4-6-20260217`.
Open WebUI Web-based chat UI	Tested	Chat completions via OpenAI connection type Tested: March 2026 — see our Open WebUI guide	Admin → Connections → OpenAI API. Paste proxy URL + key. Select model from dropdown.
LangChain (Python) langchain-openai	Tested	ChatOpenAI with memory chains Tested: March 2026	`from langchain_openai import ChatOpenAI llm = ChatOpenAI( openai_api_base="https://cheapai-netifly-app.up.railway.app/v1", openai_api_key="YOUR_KEY", model="claude-sonnet-4-6-20260217" )`
LlamaIndex llama-index-llms-openai	Partially Tested	Basic completions confirmed. RAG pipelines not yet validated. Tested: February 2026	`from llama_index.llms.openai import OpenAI llm = OpenAI( api_base="https://cheapai-netifly-app.up.railway.app/v1", api_key="YOUR_KEY" )`
Windsurf IDE Custom OpenAI endpoint	Partially Tested	Chat completions work. Native Cascade mode not yet tested. Tested: March 2026	Set custom OpenAI base URL in Settings. Function calling not yet systematically validated.
n8n AI Nodes OpenAI credential type	Not Yet Tested	Expected to work via OpenAI credential with custom base URL. Formal test pending	Test pending — no reproducible evidence yet.
Custom Serverless / Vercel Edge fetch / axios REST	Tested	Plain fetch against /v1/chat/completions. No IP blocking. Tested: March 2026	`const res = await fetch( 'https://cheapai-netifly-app.up.railway.app/v1/chat/completions', { method:'POST', headers:{Authorization:'Bearer KEY'}, body: JSON.stringify({model:'gpt-5.4', messages:[...]}) } );`

Testing was performed by the CheapAI team against the live endpoint. If you have tested an unlisted integration, contact us via Telegram or Discord to get it added.