Client Apps
No code needed. Any app that accepts a custom OpenAI endpoint can run on AkiliNode — you only ever enter the same three values:
| Setting | Value |
|---|---|
| Provider type | OpenAI (or "OpenAI-compatible") |
| Base URL / API host | https://api.akilinode.com/v1 |
| API key | Your AkiliNode token (sk-...) |
| Model | deepseek-flash or deepseek-v4-pro |
Cherry Studio
- Open Settings → Model Providers and add a custom provider (or edit the OpenAI provider).
- Set the API host to
https://api.akilinode.comand paste your token as the API key. - Add the models
deepseek-flashanddeepseek-v4-pro, then pick one in a new chat.
ChatBox
- Open Settings → Model and choose to add a custom provider.
- API mode: OpenAI API. API host:
https://api.akilinode.com/v1. - Paste your token, set the model to
deepseek-flash, and save.
Lobe Chat / other web clients
- In the model settings, select the OpenAI provider.
- Override the endpoint with
https://api.akilinode.com/v1and enter your token. - If the client asks for a model list, add the two model names manually.
Coding tools (Cursor, Continue, and similar)
Most AI coding tools let you bring your own OpenAI-compatible key. Look for "OpenAI API Key / Override base URL" in settings, enter your AkiliNode token and https://api.akilinode.com/v1, and choose one of the two models. Exact menu names differ between versions — the three values above are always the same.
Tip: Create a separate token for each app in the console and give each one a spending cap. If a key ever leaks, you can disable just that token without touching your other apps.
Trouble connecting?
- 401 / invalid key — the token was copied with a space, or the token is disabled. Create a fresh token and paste carefully.
- Model not found — check the spelling:
deepseek-flash,deepseek-v4-pro. - Balance error — the app works but your credit is finished. See Account & Billing.
- More fixes in Help & Support.
© 2026 AkiliNode · Frontier AI APIs, priced for Africa