Connect with the SDK
Choose your model.
Hoveydai is an OpenAI-compatible API. You change the base URL and the key, pick the model with a string on every request, and see the cost of each key separately.
- Base URL and keySet these alongside a supported model ID; check endpoint and parameter compatibility.
- ModelOne string per request; the library stays the same.
- CostUsage is tracked per key against shared account credit.
from openai import OpenAI -client = OpenAI(api_key="sk-...")+client = OpenAI(+ base_url="https://<HOVEYDAI_BASE_URL>/v1",+ api_key="hvd_...",+) resp = client.chat.completions.create( model="gpt-5.5", # any model id from the list messages=[{"role": "user", "content": "..."}],)
From key to first reply in three steps.
Create a key
In the dashboard, create a separate key for each app so its cost is shown separately.
Change the base URL
Set the base URL and the key in the SDK settings. Use a model ID from the catalog and review supported parameters.
Send a request
Write the model name and send. The reply comes back in the OpenAI format and the cost is recorded against that key.
Any library that knows OpenAI.
Works with the official OpenAI SDK in Python and JavaScript, streaming or not. Compatibility applies to the documented endpoints and parameters; provider-specific SDK features may differ.
from openai import OpenAI client = OpenAI( base_url="https://<HOVEYDAI_BASE_URL>/v1",changed api_key="hvd_...",) resp = client.chat.completions.create( model="gpt-5.5", messages=[{"role": "user", "content": "hello"}],)print(resp.choices[0].message.content)
Different models. One connection.
Change the model in your request. Manage your API keys, usage reports and account credit in the same platform.
- One key for allChange the model name in the request, not the library.
- Shared account creditChat and API usage draw from the same account.
- One price tableInput and output price of every model, visible before you use it.
base_url="https://<HOVEYDAI_BASE_URL>/v1"Choose a model. Understand the cost.
Pick a model to update the estimator below. Sample prices are clearly labelled when the live catalog is unavailable.
Estimate monthly cost with the selected model
Sample assumption: about 800 input and 600 output tokens per message, a 30-day month. Prices are samples.
When you hit a limit, the answer is clear.
Errors come in the OpenAI error format and each one says what happened and what to do. The final list of codes is in the docs.
| Code | Means | What to do |
|---|---|---|
| 400 | Invalid request | Check the body and parameters against the docs. |
| 401 | Invalid or missing key | Check the Authorization header and the key. |
| 402 | Not enough credit | Review account credit and plan limits; see the HTTP 402 errors in the docs. |
| 404 | Model not found | Take the model id from the model list. |
| 429 | Over the request limit | Wait a moment and send again; check your plan's limit. |
| 5xx | Error on our side or the provider's | Try again in a few seconds. |
- Active planDaily and monthly requests See dashboard and docs
- Each requestContext length See dashboard and docs
- APIRate and concurrency See dashboard and docs
{ "error": { "message": "Invalid API key", "type": "invalid_request_error" }}
One key per app. Clear usage per key.
See how much each key used and what it cost. Compare the support bot with the test script.
- Usage and costTokens and USD for each key, separately.
- Delete is archiveAn archived key stops working but its history stays.
| Name | Key | Tokens this month | Cost (USD) | Status |
|---|---|---|---|---|
| Support botCreated Aug 20, 2026 | hvd_…a41f | 48,200,000 | 23.1 | Active |
| Mobile appCreated Sep 2, 2026 | hvd_…9c02 | 12,600,000 | 7.4 | Active |
| Test scriptCreated Sep 26, 2026 | hvd_…5b77 | 310,000 | 0.15 | Archived |