Install ChatBox first
Install ChatBox or open the ChatBox web app first. Then use the settings button in the sidebar, open Model Provider, add a custom provider if CorvusLLM is not already listed, and choose the OpenAI API compatible provider type.
| What to install | Official source | Quick check |
|---|---|---|
| ChatBox app or web app | ChatBox download | Open the app before editing provider settings. |
| Custom provider settings | ChatBox provider docs | Settings sidebar > Model Provider > Add provider. |
| CorvusLLM key | Buy or top up balance | Paste as the API key. |
Zero-to-working setup
Use this block when you want the shortest reliable path from a clean install to one successful ChatBox request through CorvusLLM.
- ChatBox desktop app or web app
- CorvusLLM key
Official source: ChatBox provider docs, ChatBox downloads
Open Settings sidebar > Model Provider > Add, then choose an OpenAI API compatible or custom provider.
- Open ChatBox settings from the sidebar
- Open Model Provider settings and add a provider if needed
- Choose OpenAI API compatible, then paste API Host and API Key
- Use ChatBox Check if your version shows it
- Create a new conversation and send Say chatbox-ok with gpt-5.5
ChatBox Check passes or a new conversation returns the short test answer with the selected model.
| Value | Paste this |
|---|---|
| API Host | https://base.corvusllm.com/v1 |
| API Key | YOUR_CORVUSLLM_KEY |
| Model | gpt-5.5 |
ChatBox usually wants the API host/base URL, not a full /chat/completions URL.
Desktop or mobile
Open ChatBox settings, add a custom provider, choose the OpenAI-compatible option, fill the API Host and API Key fields, then add at least one model slug. The exact labels can vary by version, but these fields are the ones to match.
| Field | Where to change it | Value |
|---|---|---|
| Provider type | Settings sidebar > Model Provider > Add | OpenAI-compatible or custom OpenAI provider |
| Base URL | API Host or custom provider base URL input |
https://base.corvusllm.com/v1
|
| API key | API Key input in the same custom provider form | Your delivered CorvusLLM key |
| Model | Model name, model ID, or custom model input |
gpt-5.5
|
Provider: OpenAI API compatible
API Host: https://base.corvusllm.com/v1
API Key: YOUR_CORVUSLLM_KEY
Model: gpt-5.5 Connection test
Save the provider, click Check if your ChatBox version shows a connection check button, open a new ChatBox conversation, select the custom model, and send a short test message. If ChatBox lets you inspect provider logs, confirm it is calling https://base.corvusllm.com/v1 and not a direct provider endpoint.
- If the provider fails immediately, recopy the API key and base URL.
- If the model field is blank, paste
gpt-5.5manually before switching to another slug. - If a normal text prompt works but images or tools fail, test the same request shape in a simpler client before changing every setting.
Important note
Use the base URL only. Do not type /chat/completions into the host field unless your specific ChatBox build explicitly asks for a full endpoint URL.