WorkBuddy
WorkBuddy is a desktop AI workbench. Add AccelsRouter as a custom model and every model AccelsRouter supports becomes available in its chat — configured entirely through the GUI, no config files.
Prerequisites
| Item | Requirement |
|---|---|
| WorkBuddy | Latest |
| AccelsRouter | A working sk- token |
Configure
Open Settings → Model → Add model, then pick the Custom access method and fill in:
| Field | Value |
|---|---|
| URL | https://router.accels.tech/v1 |
| API Key | sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx |
| Model name | An exact model id, e.g. claude-sonnet-4-6 |
Save — the configuration persists automatically. WorkBuddy uses the standard /chat/completions path by default, appending it to the URL above.
If you point at a gateway with a non-standard path, turn on the Custom protocol switch in advanced settings; WorkBuddy then uses the address exactly as entered. For AccelsRouter, leave it off and use
https://router.accels.tech/v1.
Verify
- Open the chat model selector and pick your AccelsRouter model.
- Send "Hello" — you should get a response.
- If a model is greyed out, its URL or API Key is missing — open its configuration and complete it.
Recommended models
| Model | Use |
|---|---|
claude-sonnet-4-6 | Daily driver |
claude-opus-4-8 | Long-form writing, deep analysis |
gpt-5.4 | Multimodal (images, PDFs) |
gemini-2.5-pro | 1M-token context |
deepseek-v4-pro | Reasoning |
Full list at https://router.accels.tech/pricing.
Known limitations
- Add one custom model at a time; repeat for each model id you want available.
- Capability flags (tool calling, image input, reasoning) are not auto-filled in Custom mode — enable the ones your chosen model supports.