Use gpt88.cc in ChatBox

The goal here is not just to mention a config field. It is to take ChatBox from a blank setup to a working state: API key, base URL, model ID, then one verification round.

Preparation

Confirm these items before you start:

1. Install the ChatBox desktop app
2. Create a gpt88.cc API key
3. Confirm the exact model ID you want to use
4. Know whether your client is using OpenAI-style or Claude-style settings
5. Prepare one minimal test prompt

Step 1: Open ChatBox and enter settings

On first launch, ChatBox usually shows a setup wizard. If it has already been configured, reopen the settings panel. The immediate target is simple: set the provider and the API host correctly.

1. Launch ChatBox.
2. If this is the first run, enter the setup wizard or create a new chat.
3. If it is already configured, open settings from the lower-left corner.
4. Find the AI Provider or model provider section.

Step 2: Configure gpt88.cc

2.1 Choose the provider type

For gpt88.cc, the most common option is OpenAI API. If your ChatBox build also exposes a Claude API option and you specifically want Claude-family models, you can switch to that route.

2.2 Fill in API settings

Recommended first pass: OpenAI-compatible settings.

Provider: OpenAI API
API Key: sk-your-gpt88-api-key
API Host / API Domain: https://api.gpt88.cc
Model: claude-haiku-4-5-20251001 or gpt-5-2-chat-latest

If you intentionally want the Claude-style route, use this instead:

If your ChatBox build exposes a Claude API option:

Provider: Claude API
API Key: sk-your-gpt88-api-key
API Host / API Domain: https://api.gpt88.cc
Model: claude-sonnet-4-6 or claude-haiku-4-5-20251001

The two most common mistakes are the host shape and the exact model name:

2.3 Select the model

After the host is set, the next critical step is the model ID. ChatBox supports manual entry, and direct pasting is often faster than waiting for a model list to refresh.

Step 3: Run the first test

After saving the config, go back to the main screen and send one minimal request. Confirm a normal response first, then tune temperature, output length, or switch models.

export GPT88_API_KEY="sk-your-gpt88-api-key"

curl https://api.gpt88.cc/v1/chat/completions \
  -H "Authorization: Bearer $GPT88_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5-2-chat-latest",
    "messages": [
      {"role": "user", "content": "Introduce gpt88.cc in one sentence."}
    ]
  }'

1. Create a new chat.
2. Send a simple request such as “Introduce gpt88.cc in one sentence.”
3. Confirm that the response returns normally.
4. Only after that should you move to larger prompts or different models.

Advanced notes

Common questions

• Can I use Claude-family models in ChatBox? Yes, as long as you use the correct host style and the exact model ID.
• Do I need the model list dropdown? No. Manual model ID entry is often the fastest route.
• Should I start with the strongest model? No. Start with a cheaper stable model so you can separate config issues from model behavior.

Usage tips

• Keep one dedicated ChatBox profile for gpt88.cc instead of mixing providers in one config.
• Copy real model IDs directly from the model catalog to avoid hidden typos.
• When troubleshooting, verify with curl first, then come back to the UI client.

Troubleshooting checklist

1. Cannot connect
   - Make sure the Base URL is https://api.gpt88.cc
   - Make sure the API key is complete
   - Verify with curl before going back to ChatBox

2. Model list is empty
   - Enter the model ID manually
   - Copy the real model name from /v1/models or the model catalog

3. 401 response
   - The key is invalid, expired, or lacks permission

4. 404 response
   - Usually the Base URL or model name is wrong

5. Slow responses
   - First test with a lighter model to separate config problems from model latency

Next steps

• Read the Chat Completions API reference.
• Browse available model IDs.
• Return to the integrations hub.