HopBase DocsWorkBuddy
Use HopBase as WorkBuddy's OpenAI-compatible model provider.
WorkBuddy can call an OpenAI-compatible endpoint as its underlying model backend, so HopBase can serve the models behind it.
Generate the config in one step (recommended)
Paste a key from a chat group. The generator uses it to call GET /v1/models, lists the models WorkBuddy can use, and builds a command that writes the config file. Model IDs must match exactly, and reading them from your key rules out typos.
The config file is ~/.workbuddy/models.json, or %USERPROFILE%\.workbuddy\models.json on Windows. WorkBuddy reloads it automatically, so the new models show up in the model dropdown right after you write it.
The command replaces the whole file and first backs up the old one as models.json.bak-<time>. If you had added models by hand, copy those entries back from the backup.
Fill in the settings by hand
| Setting | Value |
|---|---|
| Protocol / provider type | OpenAI (Chat Completions) |
| API address / Base URL | https://api.hop-base.com/v1 |
| API key | A key for a chat model group |
| Model | The exact ID returned by GET /v1/models for that key |
The settings screen takes the base URL. If you edit models.json directly, url needs the full endpoint:
https://api.hop-base.com/v1/chat/completionsVerified in production: WorkBuddy reaches POST /v1/chat/completions and streams normally.
Pick a chat-capable key
Sign in to the console and use a key whose group serves OpenAI-protocol chat models — GPT / Codex, GLM, Kimi K3, DeepSeek, Qwen, or Grok.
Read the exact model ID
curl https://api.hop-base.com/v1/models \
-H "Authorization: Bearer sk-your-key"Copy an ID from the response verbatim.
Fill in the provider
In WorkBuddy's model provider settings, choose the OpenAI-compatible type, then enter the base URL, the key, and the model ID from the steps above.
Send a test message
Ask for a one-word reply. A successful round trip confirms the key, the group, the protocol, and the model ID all line up.
Common errors
| Error | Cause | Fix |
|---|---|---|
route_not_found | The key belongs to an image or video group with no chat endpoint | Use a key from a chat model group |
404 model_not_found | The model ID is a short name such as glm | Use the full ID, such as glm-5.3 |
Groups such as Kling, Seedance, MiniMax, and Wan / HappyHorse serve task-based media endpoints, not chat/completions. Model IDs carry a version or date, so always copy them from GET /v1/models.
Claude models
Claude uses the Anthropic protocol, not Chat Completions, so WorkBuddy cannot call it. Point WorkBuddy at a chat group that serves OpenAI-protocol models. See Third-party clients for the protocol-per-family table.
Long system prompts
WorkBuddy sends a large instruction context, so a single turn can carry tens of thousands of input tokens. Pick a model whose context window has room for it, and be aware that some families change pricing tier once input crosses a threshold — see the family's own page under Model integration.
Next steps
- Third-party clients — the same settings for Cherry Studio, NextChat, and LobeChat
- Available models — what each plan group serves
- Troubleshooting — look up an error message