# TypingMind setup

TypingMind is a chat front-end that runs in the browser. Add each Claudech model as a custom model.

## Before you start

- **A Claudech API key** from [API keys](https://claudech.com/api-keys). API access unlocks after your first purchase.

## Add a custom model

Open **Models → Add Custom Model** and fill in:

| Field | Value |
| --- | --- |
| Name | `Claude Opus 5.5` (any label) |
| API type | **OpenAI compatible** |
| Endpoint | `https://api.claudech.com/v1/chat/completions` |
| Model ID | `claude-opus-5-5` |
| Context length | `1000000` |
| Authentication | `Authorization: Bearer sk-...` |

Turn on image and tool support if your version offers them; every Claudech model supports both. Use **Test** to check the connection, then save. Repeat for each model you want.

> **Full endpoint here**
>
> Unlike most tools, TypingMind asks for the full endpoint including `/chat/completions`, not just the base URL.

## Troubleshooting

| Symptom | Fix |
| --- | --- |
| Test fails with `401` | Check the header is `Authorization` with the value `Bearer sk-...` |
| Test fails with `404` | The endpoint must end in `/v1/chat/completions` |
| `402 insufficient_balance` | Top up under [Billing](https://claudech.com/billing) |

See [all integrations](https://claudech.com/docs/integrations) for other chat apps. Claudech is an independent service and is not affiliated with the makers of TypingMind.

---

Source: https://claudech.com/docs/integrations/typingmind