Integration guide

Connect your account, create an AI API, then add the URL, key, model and protocol configuration to your tools or code.

BEFORE YOU CONNECT

Understand the connection before you pay.

You bring the AI account. We host the connection, manage separate keys and help you inspect requests. Registration is free; connecting a real account requires an active paid plan.

Why authorize an account?: The platform needs credentials to send requests to your provider on behalf of your apps. Subscription sign-in happens on the provider’s official page; API connections use the provider key you supply. Your apps use separate platform-generated keys.
What is stored?: We store account and plan information, connection settings, encrypted authorization credentials and recoverable API keys, plus request status, model, usage, timing and errors. Request body storage is controlled in account settings; when enabled, it is encrypted and subject to retention limits.
How do I stop access?: Pause a connection to stop new requests while retaining credentials. Delete it to clear credentials held by the platform and schedule related data cleanup. To revoke provider-side access, also revoke the connection with the provider where supported, or revoke or rotate your provider API key.
Platform fees and provider limits: Platform fees cover the hosted gateway, connections, app keys and request history, within your plan’s connection, concurrency and retention limits. They exclude provider subscriptions and API usage, do not increase or reset provider allowances, and do not bypass rate limits. Check that your account, model and intended use are supported first.

1. Connect an account

Register, verify your email and activate a plan. In My access, choose Add connection and select an available connection product.

Codex subscription: name the connection and get an authorization code. Read the instructions, open the official sign-in page, sign in to your own subscription account, and enter the code. Return to wait for authorization to finish.
API products: enter the API key issued by that provider, then save and check the connection.

Once the account is connected, create an AI API yourself; connecting does not automatically create a key. Pausing keeps its authorization and stops new requests until you enable it again.

Open My access →

2. Create your AI API

In My access, choose Create my AI API and select the account and models. Available models are selected by default. You can change the name; leaving expiry blank means no expiry.

Create separate AI APIs for different apps. Each key can call only its selected models. Expiry takes effect at the end of the selected day in your account time zone. Choose Use API after creation to get the configuration.

New keys support Show and Copy. Legacy keys whose full values were not stored cannot be recovered; create a replacement. Deleting a key requires typing DELETE. The API record and key are permanently removed; historical requests and usage are retained.

Open my AI APIs →

3. Configure your app

Base URL / API URL: Paste the Base URL shown in Use API, normally https://gateway.llmtoapi.com/v1. Do not append /v1 twice.
API Key: Use the selected AI API key, not your provider account credentials.
Model: Choose Use API on the selected row, select a model and protocol, and copy the actual model ID. Do not construct or alter model names yourself.

Use API lists the selected model’s supported protocols and provides cURL, Python SDK, environment and compatible client configuration. Codex uses Responses. Messages requires a compatible client and cannot use Chat configuration.

Get my integration settings →

Request examples

Replace YOUR_API_KEY and YOUR_MODEL with your settings, or copy a complete example for the current model and protocol from Use API.

Chat Completions

curl 'https://gateway.llmtoapi.com/v1/chat/completions' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"model":"YOUR_MODEL","messages":[{"role":"user","content":"Hello"}]}'

Responses

curl 'https://gateway.llmtoapi.com/v1/responses' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"model":"YOUR_MODEL","input":"Hello","stream":true}'

These are shell/curl examples. Use the Playground to test the platform request, then send a request from your own client and check the request logs. Real requests consume your provider account allowance.

Messages

curl 'https://gateway.llmtoapi.com/v1/messages' \
  -H 'x-api-key: YOUR_API_KEY' \
  -H 'anthropic-version: 2023-06-01' \
  -H 'Content-Type: application/json' \
  -d '{"model":"YOUR_MODEL","max_tokens":1024,"messages":[{"role":"user","content":"Hello"}]}'

Use this example only when the selected model lists Messages. Codex Chat compatibility has limitations for tool calls and image input; choose a supported protocol and test the intended feature.

Troubleshooting

401: check whether the API key is correct, expired, disabled or deleted.
403: check allowed models, whether the key matches the proxy URL, and provider access under your plan.
Connection issue: use the recovery action in Getting started or My access. Update the original authorization, adjust an expired key’s expiry, or enable a paused key or connection. Creating a new key does not restore authorization.
429: reduce concurrency or request frequency and check your provider allowance.

If the issue continues, inspect the error in Request logs and contact support from your account menu. Do not include full API keys in tickets.