Configure Chatbox
Chatbox is a cross-platform AI client for Windows, macOS, Linux, iOS, and Android. Its custom model-provider feature can connect LAYSO AI through OpenAI, Responses, Anthropic, or Gemini-compatible modes.
One key for every platform model
One LAYSO AI API key can access every model available on the platform. Billing follows the selected model's current price. Every successful request has an itemized charge, while failed tasks are not charged.
Prerequisites
- Create and copy a key on the API Keys page.
- Choose a model on Models and pricing, then copy the exact model ID displayed there.
- Install the correct version for your device from the Chatbox website.
Do not guess model names
Use the model ID, context length, and capabilities shown on the Models and pricing page. Do not remove provider prefixes or copy arbitrary example model names.
Protocols and addresses
Chatbox lets a custom provider use different API modes. Match the host to the protocol supported by the target model:
| Chatbox API mode | API Host | Use case |
|---|---|---|
| OpenAI API Compatible | https://api.layso.ai/v1 | Chat Completions, tool calls, and vision |
| OpenAI Responses API Compatible | https://api.layso.ai/v1 | Responses API models and new projects |
| Anthropic Claude API Compatible | https://api.layso.ai | Native-compatible Anthropic Messages |
| Google Gemini API Compatible | https://api.layso.ai | Native-compatible Gemini requests |
Address rule
The two OpenAI modes use https://api.layso.ai/v1. Anthropic and Gemini modes use the gateway root https://api.layso.ai. Do not append another /v1, /messages, /v1beta, or a complete request path.
Desktop setup
Step 1: Open Settings
Launch Chatbox and select Settings in the lower-left sidebar.
/screenshots/chatbox/01-open-settings.pngSuggested screenshot: highlight the desktop Settings entry.Step 2: Add a model provider
Select Model Provider in the settings sidebar, then select + Add Provider at the bottom of the list.
/screenshots/chatbox/02-add-provider.pngSuggested screenshot: highlight Model Provider and + Add Provider.Step 3: Select an API mode
Enter LAYSO AI as the name, then choose a protocol for the target model:
- Use OpenAI API Compatible for typical chats.
- Use OpenAI Responses API Compatible when the model and workflow require Responses.
- Use Anthropic Claude API Compatible for the native Anthropic request format.
- Use Google Gemini API Compatible for the native Gemini request format.
Select Add after choosing the protocol.
/screenshots/chatbox/03-select-api-mode.pngSuggested screenshot: show the provider name and all four API modes.Step 4: Enter connection details
Fill in the fields for the selected API mode:
| Chatbox field | Value |
|---|---|
| Name | LAYSO AI |
| API Key | The complete key copied from the API Keys page |
| API Host | The matching address from the protocol table above |
Select Fetch to load the model list. Paste the API key only into the key field, never into Name or Model.
/screenshots/chatbox/04-provider-fields.pngSuggested screenshot: show API Key, API Host, and Fetch; mask the key.Step 5: Enable a model
In the fetched list, select the + next to a model to add it to the enabled list.
If the list is empty or the target model is missing:
- Open Models and pricing.
- Copy the exact model ID.
- Return to Chatbox and add the model manually.
- Set the model type to Chat.
- Enable vision, reasoning, or tool-use capabilities only when the model page lists them.
/screenshots/chatbox/05-enable-model.pngSuggested screenshot: show the model list, + button, and capability options.Step 6: Check the connection
Select Check, then choose an enabled model for the test.
Save after Chatbox reports a successful connection. If the dialog checks text, images, and tool calls separately, only capabilities supported by the selected model need to pass.
/screenshots/chatbox/06-connection-success.pngSuggested screenshot: show the test model and successful result.Mobile setup
Chatbox for iOS and Android uses the same LAYSO AI addresses, although navigation differs from the desktop app.
Step 1: Open Settings and add a provider
Open the sidebar from the upper-left menu, select Settings, then select Add Model Provider.
Enter:
- Name:
LAYSO AI - API mode: choose from the protocol table above
- API Key: copy it from the API Keys page
- API Host: use the matching address from the protocol table
/screenshots/chatbox/07-mobile-provider.pngSuggested screenshot: show Name, API mode, API Key, and API Host on mobile.Step 2: Configure model parameters
After saving the provider:
- Copy the exact model ID from Models and pricing.
- Set the model type to Chat.
- Enable vision, reasoning, or tool-use only when the model supports it.
- Save the model.
Step 3: Test the connection
Select Test Model. Return to the home screen after it succeeds. If it fails, use the troubleshooting checklist below.
/screenshots/chatbox/08-mobile-test.pngSuggested screenshot: show a successful mobile model test.Start using Chatbox
- Close Settings and return to the Chatbox home screen.
- Open the model selector near the prompt box.
- Select the enabled model under
LAYSO AI. - Send a short message and confirm that a response returns.
/screenshots/chatbox/09-first-chat.pngSuggested screenshot: show the LAYSO AI group, selected model, and first response.Verification and billing
After the first successful chat, confirm that:
- The active Chatbox provider is
LAYSO AI. - The response model matches the selected exact model ID.
- The successful request has an itemized charge in Usage logs.
- A failed request did not create a charge.
- No complete API key appears in screenshots, logs, or exported settings.
Troubleshooting
Fetch returns an empty model list
- Confirm that API Host matches the selected API mode.
- Remove an unnecessary trailing slash and check for duplicated protocol paths.
- Copy the API key again without spaces or line breaks.
- If fetching still fails, add the exact model ID manually from Models and pricing.
Connection check fails
- Confirm that the current network can access
https://api.layso.ai. - Confirm that the API key is still active.
- Confirm that the model ID supports the selected API mode.
- Save, fully quit Chatbox, and reopen it.
401 or 403
Copy the key again from the API Keys page. Do not use another platform's key or manually include the Bearer prefix in Chatbox's API Key field.
404
This usually means that API mode and API Host do not match:
- OpenAI and Responses:
https://api.layso.ai/v1 - Anthropic and Gemini:
https://api.layso.ai
Do not enter a complete endpoint such as /chat/completions, /responses, /messages, or /v1beta/models in API Host.
Text works, but vision or tool-use checks fail
The model may not support that capability, or its capability toggle may be incorrect. Follow the capability information on Models and pricing.
Privacy and security
- Never share a complete API key in screenshots, chats, logs, or support tickets.
- Remove the custom provider and key after using a shared device.
- If a key may be exposed, delete it on the API Keys page and create a replacement immediately.
- Chatbox is a third-party client; refer to Chatbox's documentation for local storage, sync, and data-handling behavior.
Replacing the image placeholders
This page reserves image positions only. Put screenshots in docs/public/screenshots/chatbox/, then replace a placeholder with standard Markdown, for example:
